Wine Classifier Notebook.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Introduction"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this project, we will analyze and classify a wine color dataset. The dataset contains various wine parameters, such as acidity, alcohol content and pH. Our goal is to test different classification algorithms on two versions of this dataset: a less processed and a more processed version."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import StandardScaler, OneHotEncoder, MinMaxScaler\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn.metrics import accuracy_score, confusion_matrix\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from scipy import stats\n",
    "\n",
    "classifiers = []\n",
    "names = []"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Database presentation"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, we will load the data from the CSV file and take a look at it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>type</th>\n",
       "      <th>fixed acidity</th>\n",
       "      <th>volatile acidity</th>\n",
       "      <th>citric acid</th>\n",
       "      <th>residual sugar</th>\n",
       "      <th>chlorides</th>\n",
       "      <th>free sulfur dioxide</th>\n",
       "      <th>total sulfur dioxide</th>\n",
       "      <th>density</th>\n",
       "      <th>pH</th>\n",
       "      <th>sulphates</th>\n",
       "      <th>alcohol</th>\n",
       "      <th>quality</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>white</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.27</td>\n",
       "      <td>0.36</td>\n",
       "      <td>20.7</td>\n",
       "      <td>0.045</td>\n",
       "      <td>45.0</td>\n",
       "      <td>170.0</td>\n",
       "      <td>1.0010</td>\n",
       "      <td>3.00</td>\n",
       "      <td>0.45</td>\n",
       "      <td>8.8</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>white</td>\n",
       "      <td>6.3</td>\n",
       "      <td>0.30</td>\n",
       "      <td>0.34</td>\n",
       "      <td>1.6</td>\n",
       "      <td>0.049</td>\n",
       "      <td>14.0</td>\n",
       "      <td>132.0</td>\n",
       "      <td>0.9940</td>\n",
       "      <td>3.30</td>\n",
       "      <td>0.49</td>\n",
       "      <td>9.5</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>white</td>\n",
       "      <td>8.1</td>\n",
       "      <td>0.28</td>\n",
       "      <td>0.40</td>\n",
       "      <td>6.9</td>\n",
       "      <td>0.050</td>\n",
       "      <td>30.0</td>\n",
       "      <td>97.0</td>\n",
       "      <td>0.9951</td>\n",
       "      <td>3.26</td>\n",
       "      <td>0.44</td>\n",
       "      <td>10.1</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>white</td>\n",
       "      <td>7.2</td>\n",
       "      <td>0.23</td>\n",
       "      <td>0.32</td>\n",
       "      <td>8.5</td>\n",
       "      <td>0.058</td>\n",
       "      <td>47.0</td>\n",
       "      <td>186.0</td>\n",
       "      <td>0.9956</td>\n",
       "      <td>3.19</td>\n",
       "      <td>0.40</td>\n",
       "      <td>9.9</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>white</td>\n",
       "      <td>7.2</td>\n",
       "      <td>0.23</td>\n",
       "      <td>0.32</td>\n",
       "      <td>8.5</td>\n",
       "      <td>0.058</td>\n",
       "      <td>47.0</td>\n",
       "      <td>186.0</td>\n",
       "      <td>0.9956</td>\n",
       "      <td>3.19</td>\n",
       "      <td>0.40</td>\n",
       "      <td>9.9</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    type  fixed acidity  volatile acidity  citric acid  residual sugar  \\\n",
       "0  white            7.0              0.27         0.36            20.7   \n",
       "1  white            6.3              0.30         0.34             1.6   \n",
       "2  white            8.1              0.28         0.40             6.9   \n",
       "3  white            7.2              0.23         0.32             8.5   \n",
       "4  white            7.2              0.23         0.32             8.5   \n",
       "\n",
       "   chlorides  free sulfur dioxide  total sulfur dioxide  density    pH  \\\n",
       "0      0.045                 45.0                 170.0   1.0010  3.00   \n",
       "1      0.049                 14.0                 132.0   0.9940  3.30   \n",
       "2      0.050                 30.0                  97.0   0.9951  3.26   \n",
       "3      0.058                 47.0                 186.0   0.9956  3.19   \n",
       "4      0.058                 47.0                 186.0   0.9956  3.19   \n",
       "\n",
       "   sulphates  alcohol  quality  \n",
       "0       0.45      8.8        6  \n",
       "1       0.49      9.5        6  \n",
       "2       0.44     10.1        6  \n",
       "3       0.40      9.9        6  \n",
       "4       0.40      9.9        6  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv(\"winequalityN.csv\")\n",
    "data.head()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data preprocessing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this section, we will check for missing values and errors in our dataset. If so, we will repair them appropriately."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Checking for missing data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "type                     0\n",
       "fixed acidity           10\n",
       "volatile acidity         8\n",
       "citric acid              3\n",
       "residual sugar           2\n",
       "chlorides                2\n",
       "free sulfur dioxide      0\n",
       "total sulfur dioxide     0\n",
       "density                  0\n",
       "pH                       9\n",
       "sulphates                4\n",
       "alcohol                  0\n",
       "quality                  0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "       type  fixed acidity  volatile acidity  citric acid  residual sugar  \\\n",
      "17    white            NaN             0.660         0.48             1.2   \n",
      "174   white            NaN             0.270         0.31            17.7   \n",
      "249   white            NaN             0.410         0.14            10.4   \n",
      "267   white            NaN             0.580         0.07             6.9   \n",
      "368   white            NaN             0.290         0.48             2.3   \n",
      "518   white            NaN             0.130         0.28             1.9   \n",
      "1079  white            NaN               NaN         0.29             6.2   \n",
      "2902  white            NaN             0.360         0.14             8.9   \n",
      "6428    red            NaN             0.440         0.09             2.2   \n",
      "6429    red            NaN             0.705         0.10             2.8   \n",
      "\n",
      "      chlorides  free sulfur dioxide  total sulfur dioxide  density    pH  \\\n",
      "17        0.029                 29.0                  75.0  0.98920  3.33   \n",
      "174       0.051                 33.0                 173.0  0.99900  3.09   \n",
      "249       0.037                 18.0                 119.0  0.99600  3.38   \n",
      "267       0.043                 34.0                 149.0  0.99440  3.34   \n",
      "368       0.049                 36.0                 178.0  0.99310  3.17   \n",
      "518       0.050                 20.0                  78.0  0.99180  3.43   \n",
      "1079      0.046                 29.0                 227.0  0.99520  3.29   \n",
      "2902      0.036                 38.0                 155.0  0.99622  3.27   \n",
      "6428      0.063                  9.0                  18.0  0.99444   NaN   \n",
      "6429      0.081                 13.0                  28.0  0.99631   NaN   \n",
      "\n",
      "      sulphates  alcohol  quality  \n",
      "17         0.39     12.8        8  \n",
      "174        0.64     10.2        5  \n",
      "249        0.45     10.0        5  \n",
      "267        0.57      9.7        5  \n",
      "368        0.64     10.6        6  \n",
      "518        0.64     10.8        6  \n",
      "1079       0.53     10.1        6  \n",
      "2902        NaN      9.4        5  \n",
      "6428       0.69     11.3        6  \n",
      "6429       0.66     10.2        5  \n"
     ]
    }
   ],
   "source": [
    "null_data = data.loc[data['fixed acidity'].isna()]\n",
    "print(null_data)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Removing errors and missing data"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will prepare two sets of data: </br>\n",
    "\"dataLess\", in which we will remove rows that are missing, </br>\n",
    "\"dataMore\", in which we will fill in the gaps with the median of the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataLess = data.copy()\n",
    "dataLess.dropna(inplace=True)\n",
    "dataMore = data.copy()\n",
    "dataMore.fillna(dataMore.median(numeric_only=True), inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>type</th>\n",
       "      <th>fixed acidity</th>\n",
       "      <th>volatile acidity</th>\n",
       "      <th>citric acid</th>\n",
       "      <th>residual sugar</th>\n",
       "      <th>chlorides</th>\n",
       "      <th>free sulfur dioxide</th>\n",
       "      <th>total sulfur dioxide</th>\n",
       "      <th>density</th>\n",
       "      <th>pH</th>\n",
       "      <th>sulphates</th>\n",
       "      <th>alcohol</th>\n",
       "      <th>quality</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>white</td>\n",
       "      <td>6.3</td>\n",
       "      <td>0.48</td>\n",
       "      <td>0.04</td>\n",
       "      <td>1.1</td>\n",
       "      <td>0.046</td>\n",
       "      <td>30.0</td>\n",
       "      <td>99.0</td>\n",
       "      <td>0.9928</td>\n",
       "      <td>3.24</td>\n",
       "      <td>0.36</td>\n",
       "      <td>9.6</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>white</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.66</td>\n",
       "      <td>0.48</td>\n",
       "      <td>1.2</td>\n",
       "      <td>0.029</td>\n",
       "      <td>29.0</td>\n",
       "      <td>75.0</td>\n",
       "      <td>0.9892</td>\n",
       "      <td>3.33</td>\n",
       "      <td>0.39</td>\n",
       "      <td>12.8</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>white</td>\n",
       "      <td>7.4</td>\n",
       "      <td>0.34</td>\n",
       "      <td>0.42</td>\n",
       "      <td>1.1</td>\n",
       "      <td>0.033</td>\n",
       "      <td>17.0</td>\n",
       "      <td>171.0</td>\n",
       "      <td>0.9917</td>\n",
       "      <td>3.12</td>\n",
       "      <td>0.53</td>\n",
       "      <td>11.3</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     type  fixed acidity  volatile acidity  citric acid  residual sugar  \\\n",
       "16  white            6.3              0.48         0.04             1.1   \n",
       "17  white            NaN              0.66         0.48             1.2   \n",
       "18  white            7.4              0.34         0.42             1.1   \n",
       "\n",
       "    chlorides  free sulfur dioxide  total sulfur dioxide  density    pH  \\\n",
       "16      0.046                 30.0                  99.0   0.9928  3.24   \n",
       "17      0.029                 29.0                  75.0   0.9892  3.33   \n",
       "18      0.033                 17.0                 171.0   0.9917  3.12   \n",
       "\n",
       "    sulphates  alcohol  quality  \n",
       "16       0.36      9.6        6  \n",
       "17       0.39     12.8        8  \n",
       "18       0.53     11.3        6  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.loc[16:18]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>type</th>\n",
       "      <th>fixed acidity</th>\n",
       "      <th>volatile acidity</th>\n",
       "      <th>citric acid</th>\n",
       "      <th>residual sugar</th>\n",
       "      <th>chlorides</th>\n",
       "      <th>free sulfur dioxide</th>\n",
       "      <th>total sulfur dioxide</th>\n",
       "      <th>density</th>\n",
       "      <th>pH</th>\n",
       "      <th>sulphates</th>\n",
       "      <th>alcohol</th>\n",
       "      <th>quality</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>white</td>\n",
       "      <td>6.3</td>\n",
       "      <td>0.48</td>\n",
       "      <td>0.04</td>\n",
       "      <td>1.1</td>\n",
       "      <td>0.046</td>\n",
       "      <td>30.0</td>\n",
       "      <td>99.0</td>\n",
       "      <td>0.9928</td>\n",
       "      <td>3.24</td>\n",
       "      <td>0.36</td>\n",
       "      <td>9.6</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>white</td>\n",
       "      <td>7.4</td>\n",
       "      <td>0.34</td>\n",
       "      <td>0.42</td>\n",
       "      <td>1.1</td>\n",
       "      <td>0.033</td>\n",
       "      <td>17.0</td>\n",
       "      <td>171.0</td>\n",
       "      <td>0.9917</td>\n",
       "      <td>3.12</td>\n",
       "      <td>0.53</td>\n",
       "      <td>11.3</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     type  fixed acidity  volatile acidity  citric acid  residual sugar  \\\n",
       "16  white            6.3              0.48         0.04             1.1   \n",
       "18  white            7.4              0.34         0.42             1.1   \n",
       "\n",
       "    chlorides  free sulfur dioxide  total sulfur dioxide  density    pH  \\\n",
       "16      0.046                 30.0                  99.0   0.9928  3.24   \n",
       "18      0.033                 17.0                 171.0   0.9917  3.12   \n",
       "\n",
       "    sulphates  alcohol  quality  \n",
       "16       0.36      9.6        6  \n",
       "18       0.53     11.3        6  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataLess.loc[16:18]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>type</th>\n",
       "      <th>fixed acidity</th>\n",
       "      <th>volatile acidity</th>\n",
       "      <th>citric acid</th>\n",
       "      <th>residual sugar</th>\n",
       "      <th>chlorides</th>\n",
       "      <th>free sulfur dioxide</th>\n",
       "      <th>total sulfur dioxide</th>\n",
       "      <th>density</th>\n",
       "      <th>pH</th>\n",
       "      <th>sulphates</th>\n",
       "      <th>alcohol</th>\n",
       "      <th>quality</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>white</td>\n",
       "      <td>6.3</td>\n",
       "      <td>0.48</td>\n",
       "      <td>0.04</td>\n",
       "      <td>1.1</td>\n",
       "      <td>0.046</td>\n",
       "      <td>30.0</td>\n",
       "      <td>99.0</td>\n",
       "      <td>0.9928</td>\n",
       "      <td>3.24</td>\n",
       "      <td>0.36</td>\n",
       "      <td>9.6</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>white</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.66</td>\n",
       "      <td>0.48</td>\n",
       "      <td>1.2</td>\n",
       "      <td>0.029</td>\n",
       "      <td>29.0</td>\n",
       "      <td>75.0</td>\n",
       "      <td>0.9892</td>\n",
       "      <td>3.33</td>\n",
       "      <td>0.39</td>\n",
       "      <td>12.8</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>white</td>\n",
       "      <td>7.4</td>\n",
       "      <td>0.34</td>\n",
       "      <td>0.42</td>\n",
       "      <td>1.1</td>\n",
       "      <td>0.033</td>\n",
       "      <td>17.0</td>\n",
       "      <td>171.0</td>\n",
       "      <td>0.9917</td>\n",
       "      <td>3.12</td>\n",
       "      <td>0.53</td>\n",
       "      <td>11.3</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     type  fixed acidity  volatile acidity  citric acid  residual sugar  \\\n",
       "16  white            6.3              0.48         0.04             1.1   \n",
       "17  white            7.0              0.66         0.48             1.2   \n",
       "18  white            7.4              0.34         0.42             1.1   \n",
       "\n",
       "    chlorides  free sulfur dioxide  total sulfur dioxide  density    pH  \\\n",
       "16      0.046                 30.0                  99.0   0.9928  3.24   \n",
       "17      0.029                 29.0                  75.0   0.9892  3.33   \n",
       "18      0.033                 17.0                 171.0   0.9917  3.12   \n",
       "\n",
       "    sulphates  alcohol  quality  \n",
       "16       0.36      9.6        6  \n",
       "17       0.39     12.8        8  \n",
       "18       0.53     11.3        6  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataMore.loc[16:18]"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data processing"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we will prepare two versions of the data: one less processed and the other more processed."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Less processed version with changed type value to 0 and 1 and data standardization:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataLess = dataLess.replace({'type': {'white': 0, 'red': 1}})\n",
    "X_less = dataLess.drop('type', axis=1)\n",
    "y_less = dataLess['type']\n",
    "\n",
    "scaler = StandardScaler()\n",
    "X_less = scaler.fit_transform(X_less)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "More processed version with changed type value to 0 and 1, standardization, PCA transformation and removal of outliers using z-score (interquartile range in this case is too restrictive and removes all rows with red wine):"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's check if the set is balanced:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Liczba wierszy z wartością 'white':  4898\n",
      "Liczba wierszy z wartością 'red':  1599\n"
     ]
    }
   ],
   "source": [
    "sns.countplot(x='type', data=dataMore)\n",
    "plt.show()\n",
    "numberOfRows = dataMore[dataMore['type'] == 'white'].shape[0]\n",
    "print(\"Number of rows with value 'white': \", numberOfRows)\n",
    "numberOfRows = dataMore[dataMore['type'] == 'red'].shape[0]\n",
    "print(\"Number of rows with value 'red': \", numberOfRows)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's remove the outliers and check again:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of rows with value 'white':  4712\n",
      "Number of rows with value 'red' 1277\n"
     ]
    }
   ],
   "source": [
    "z_scores = np.abs(stats.zscore(dataMore.drop(columns=['type'])))\n",
    "threshold = 3  # threshold for outliers\n",
    "dataMore = dataMore[(z_scores < threshold).all(axis=1)]\n",
    "\n",
    "sns.countplot(x='type', data=dataMore)\n",
    "plt.show()\n",
    "numberOfRows = dataMore[dataMore['type'] == 'white'].shape[0]\n",
    "print(\"Number of rows with value 'white': \", numberOfRows)\n",
    "numberOfRows = dataMore[dataMore['type'] == 'red'].shape[0]\n",
    "print(\"Number of rows with value 'red'\", numberOfRows)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There is a disproportion in the data, but it is mildly unbalanced, so we will not equalize the sample sizes.\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Standardization and PCA:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataMore = dataMore.replace({'type': {'white': 0, 'red': 1}})\n",
    "\n",
    "X_more = dataMore.drop('type', axis=1)\n",
    "y_more = dataMore['type']\n",
    "\n",
    "scaler = StandardScaler()\n",
    "X_more = scaler.fit_transform(X_more)\n",
    "\n",
    "pca = PCA(n_components=2)\n",
    "X_more = pca.fit_transform(X_more)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Division into training and test set."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we will split our dataset into training and testing sets in an 80/20 ratio. About 1300/1200 rows should be enough for us to test the models well."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_less, X_test_less, y_train_less, y_test_less = train_test_split(X_less, y_less, test_size=0.2, random_state=269402)\n",
    "X_train_more, X_test_more, y_train_more, y_test_more = train_test_split(X_more, y_more, test_size=0.2, random_state=269402)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Classification"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this section, we will test different classification algorithms on both versions of our dataset. For each classifier we will calculate accuracy and create an error matrix."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Decision tree"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A less processed version"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Decision tree with pruned branches\n",
    "dt_less_pruned = DecisionTreeClassifier(max_depth=3, random_state=269402)\n",
    "dt_less_pruned.fit(X_train_less, y_train_less)\n",
    "classifiers.append(dt_less_pruned)\n",
    "names.append(f\"Decision tree with pruned branches (less)\")\n",
    "\n",
    "\n",
    "# Larger decision tree (no depth limit)\n",
    "dt_less_large = DecisionTreeClassifier(random_state=269402)\n",
    "dt_less_large.fit(X_train_less, y_train_less)\n",
    "classifiers.append(dt_less_large)\n",
    "names.append(f\"Decision tree without limits (less)\");"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A more processed version"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Decision tree with pruned branches\n",
    "dt_more_pruned = DecisionTreeClassifier(max_depth=3, random_state=269402)\n",
    "dt_more_pruned.fit(X_train_more, y_train_more)\n",
    "classifiers.append(dt_more_pruned)\n",
    "names.append(f\"Decision tree with pruned branches (more)\")\n",
    "\n",
    "# Larger decision tree (no depth limit)\n",
    "dt_more_large = DecisionTreeClassifier(random_state=269402)\n",
    "dt_more_large.fit(X_train_more, y_train_more)\n",
    "classifiers.append(dt_more_large)\n",
    "names.append(f\"Decision tree without limits (more)\");"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Naive Bayes"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A less processed version"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "nb_less = GaussianNB()\n",
    "nb_less.fit(X_train_less, y_train_less)\n",
    "classifiers.append(nb_less)\n",
    "names.append(f\"Naive Bayes (less)\");"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A more processed version"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "nb_more = GaussianNB()\n",
    "nb_more.fit(X_train_more, y_train_more)\n",
    "classifiers.append(nb_more)\n",
    "names.append(f\"Naive Bayes (more)\");"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K-Nearest Neighbors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "k_values = [1, 3, 5, 7, 9]"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A less processed version"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "for k in k_values:\n",
    "    knn_clf = KNeighborsClassifier(n_neighbors=k)\n",
    "    knn_clf.fit(X_train_less, y_train_less)\n",
    "    classifiers.append(knn_clf)\n",
    "    names.append(f'KNN k={k} (less)')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A more processed version"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "for k in k_values:\n",
    "    knn_clf = KNeighborsClassifier(n_neighbors=k)\n",
    "    knn_clf.fit(X_train_more, y_train_more)\n",
    "    classifiers.append(knn_clf)\n",
    "    names.append(f'KNN k={k} (more)')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Neural network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Various topologies and learning configurations\n",
    "hidden_layer_sizes = [(50,), (100,), (150,)]\n",
    "activations = ['identity', 'logistic', 'tanh', 'relu']\n",
    "solvers = ['lbfgs', 'sgd', 'adam']\n",
    "alphas = [0.0001, 0.001, 0.01, 0.1]"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A less processed version"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "for hidden_layer_size in hidden_layer_sizes:\n",
    "    for activation in activations:\n",
    "        for solver in solvers:\n",
    "            for alpha in alphas:\n",
    "                mlp = MLPClassifier(hidden_layer_sizes=hidden_layer_size, activation=activation,\n",
    "                                    solver=solver, alpha=alpha, max_iter=500, random_state=269402)\n",
    "                mlp.fit(X_train_less, y_train_less)\n",
    "                classifiers.append(mlp)\n",
    "                names.append(f\"MLP ({hidden_layer_size}, {activation}, {solver}, {alpha}) (less)\")\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A more processed version"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "for hidden_layer_size in hidden_layer_sizes:\n",
    "    for activation in activations:\n",
    "        for solver in solvers:\n",
    "            for alpha in alphas:\n",
    "                mlp = MLPClassifier(hidden_layer_sizes=hidden_layer_size, activation=activation,\n",
    "                                    solver=solver, alpha=alpha, max_iter=15000, random_state=269402)\n",
    "                mlp.fit(X_train_more, y_train_more)\n",
    "                classifiers.append(mlp)\n",
    "                names.append(f\"MLP ({hidden_layer_size}, {activation}, {solver}, {alpha}) (more)\")\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Summary"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this section, we will compare the results obtained for different classifiers on both versions of the data. We will determine which of the classifiers achieved the best results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "304"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(classifiers)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We created multiple classifiers, so let's see which ones had the highest accuracy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "results = {}\n",
    "for clf, name in zip(classifiers, names):\n",
    "    if \"less\" in name:\n",
    "        y_pred = clf.predict(X_test_less)\n",
    "        accuracy = accuracy_score(y_test_less, y_pred)\n",
    "    elif \"more\" in name:\n",
    "        y_pred = clf.predict(X_test_more)\n",
    "        accuracy = accuracy_score(y_test_more, y_pred)\n",
    "    results[name] = accuracy;"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's analyze more processed data separately."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KNN k=7 (more): 0.9816\n",
      "MLP ((100,), logistic, lbfgs, 0.1) (more): 0.9816\n",
      "MLP ((150,), identity, adam, 0.0001) (more): 0.9816\n",
      "MLP ((150,), identity, adam, 0.001) (more): 0.9816\n",
      "MLP ((150,), identity, adam, 0.01) (more): 0.9816\n",
      "MLP ((150,), identity, adam, 0.1) (more): 0.9816\n",
      "MLP ((150,), logistic, lbfgs, 0.001) (more): 0.9816\n",
      "MLP ((150,), relu, adam, 0.0001) (more): 0.9816\n",
      "MLP ((150,), relu, adam, 0.001) (more): 0.9816\n",
      "MLP ((150,), relu, adam, 0.01) (more): 0.9816\n",
      "KNN k=9 (more): 0.9808\n",
      "MLP ((50,), identity, lbfgs, 0.0001) (more): 0.9808\n",
      "MLP ((50,), identity, lbfgs, 0.001) (more): 0.9808\n",
      "MLP ((50,), identity, lbfgs, 0.01) (more): 0.9808\n",
      "MLP ((50,), identity, lbfgs, 0.1) (more): 0.9808\n",
      "MLP ((50,), identity, adam, 0.1) (more): 0.9808\n",
      "MLP ((50,), logistic, lbfgs, 0.001) (more): 0.9808\n",
      "MLP ((50,), logistic, adam, 0.1) (more): 0.9808\n",
      "MLP ((50,), tanh, sgd, 0.0001) (more): 0.9808\n",
      "MLP ((50,), tanh, sgd, 0.001) (more): 0.9808\n",
      "MLP ((50,), tanh, sgd, 0.01) (more): 0.9808\n",
      "MLP ((50,), tanh, sgd, 0.1) (more): 0.9808\n",
      "MLP ((50,), tanh, adam, 0.1) (more): 0.9808\n",
      "MLP ((50,), relu, lbfgs, 0.001) (more): 0.9808\n",
      "MLP ((50,), relu, adam, 0.0001) (more): 0.9808\n",
      "MLP ((50,), relu, adam, 0.001) (more): 0.9808\n",
      "MLP ((50,), relu, adam, 0.01) (more): 0.9808\n",
      "MLP ((50,), relu, adam, 0.1) (more): 0.9808\n",
      "MLP ((100,), identity, lbfgs, 0.0001) (more): 0.9808\n",
      "MLP ((100,), identity, lbfgs, 0.001) (more): 0.9808\n",
      "MLP ((100,), identity, lbfgs, 0.01) (more): 0.9808\n",
      "MLP ((100,), identity, lbfgs, 0.1) (more): 0.9808\n",
      "MLP ((150,), identity, lbfgs, 0.0001) (more): 0.9808\n",
      "MLP ((150,), identity, lbfgs, 0.001) (more): 0.9808\n",
      "MLP ((150,), identity, lbfgs, 0.01) (more): 0.9808\n",
      "MLP ((150,), identity, lbfgs, 0.1) (more): 0.9808\n",
      "MLP ((150,), logistic, lbfgs, 0.01) (more): 0.9808\n",
      "MLP ((150,), logistic, lbfgs, 0.1) (more): 0.9808\n",
      "MLP ((150,), logistic, adam, 0.0001) (more): 0.9808\n",
      "MLP ((150,), logistic, adam, 0.001) (more): 0.9808\n",
      "MLP ((150,), logistic, adam, 0.01) (more): 0.9808\n",
      "MLP ((150,), logistic, adam, 0.1) (more): 0.9808\n",
      "MLP ((150,), tanh, sgd, 0.0001) (more): 0.9808\n",
      "MLP ((150,), tanh, sgd, 0.001) (more): 0.9808\n",
      "MLP ((150,), tanh, sgd, 0.01) (more): 0.9808\n",
      "MLP ((150,), tanh, sgd, 0.1) (more): 0.9808\n",
      "MLP ((150,), tanh, adam, 0.0001) (more): 0.9808\n",
      "MLP ((150,), tanh, adam, 0.001) (more): 0.9808\n",
      "MLP ((150,), tanh, adam, 0.01) (more): 0.9808\n",
      "MLP ((150,), tanh, adam, 0.1) (more): 0.9808\n",
      "MLP ((150,), relu, adam, 0.1) (more): 0.9808\n",
      "MLP ((50,), identity, sgd, 0.0001) (more): 0.9800\n",
      "MLP ((50,), identity, sgd, 0.001) (more): 0.9800\n",
      "MLP ((50,), identity, sgd, 0.01) (more): 0.9800\n",
      "MLP ((50,), identity, sgd, 0.1) (more): 0.9800\n",
      "MLP ((50,), logistic, adam, 0.0001) (more): 0.9800\n",
      "MLP ((50,), logistic, adam, 0.001) (more): 0.9800\n",
      "MLP ((50,), logistic, adam, 0.01) (more): 0.9800\n",
      "MLP ((50,), tanh, adam, 0.0001) (more): 0.9800\n",
      "MLP ((50,), tanh, adam, 0.001) (more): 0.9800\n",
      "MLP ((50,), tanh, adam, 0.01) (more): 0.9800\n",
      "MLP ((100,), identity, sgd, 0.0001) (more): 0.9800\n",
      "MLP ((100,), identity, sgd, 0.001) (more): 0.9800\n",
      "MLP ((100,), identity, sgd, 0.01) (more): 0.9800\n",
      "MLP ((100,), identity, sgd, 0.1) (more): 0.9800\n",
      "MLP ((100,), identity, adam, 0.0001) (more): 0.9800\n",
      "MLP ((100,), identity, adam, 0.001) (more): 0.9800\n",
      "MLP ((100,), identity, adam, 0.01) (more): 0.9800\n",
      "MLP ((100,), identity, adam, 0.1) (more): 0.9800\n",
      "MLP ((100,), logistic, adam, 0.1) (more): 0.9800\n",
      "MLP ((100,), tanh, sgd, 0.0001) (more): 0.9800\n",
      "MLP ((100,), tanh, sgd, 0.001) (more): 0.9800\n",
      "MLP ((100,), tanh, sgd, 0.01) (more): 0.9800\n",
      "MLP ((100,), tanh, sgd, 0.1) (more): 0.9800\n",
      "MLP ((100,), tanh, adam, 0.0001) (more): 0.9800\n",
      "MLP ((100,), tanh, adam, 0.001) (more): 0.9800\n",
      "MLP ((100,), tanh, adam, 0.01) (more): 0.9800\n",
      "MLP ((100,), tanh, adam, 0.1) (more): 0.9800\n",
      "MLP ((100,), relu, adam, 0.0001) (more): 0.9800\n",
      "MLP ((100,), relu, adam, 0.001) (more): 0.9800\n",
      "MLP ((100,), relu, adam, 0.01) (more): 0.9800\n",
      "MLP ((100,), relu, adam, 0.1) (more): 0.9800\n",
      "MLP ((150,), identity, sgd, 0.0001) (more): 0.9800\n",
      "MLP ((150,), identity, sgd, 0.001) (more): 0.9800\n",
      "MLP ((150,), identity, sgd, 0.01) (more): 0.9800\n",
      "MLP ((150,), identity, sgd, 0.1) (more): 0.9800\n",
      "MLP ((150,), logistic, sgd, 0.0001) (more): 0.9800\n",
      "MLP ((150,), logistic, sgd, 0.001) (more): 0.9800\n",
      "MLP ((150,), logistic, sgd, 0.01) (more): 0.9800\n",
      "MLP ((150,), relu, sgd, 0.0001) (more): 0.9800\n",
      "MLP ((150,), relu, sgd, 0.001) (more): 0.9800\n",
      "MLP ((150,), relu, sgd, 0.01) (more): 0.9800\n",
      "MLP ((150,), relu, sgd, 0.1) (more): 0.9800\n",
      "MLP ((50,), identity, adam, 0.0001) (more): 0.9791\n",
      "MLP ((50,), identity, adam, 0.001) (more): 0.9791\n",
      "MLP ((50,), identity, adam, 0.01) (more): 0.9791\n",
      "MLP ((50,), logistic, lbfgs, 0.1) (more): 0.9791\n",
      "MLP ((50,), relu, sgd, 0.0001) (more): 0.9791\n",
      "MLP ((50,), relu, sgd, 0.001) (more): 0.9791\n",
      "MLP ((50,), relu, sgd, 0.01) (more): 0.9791\n",
      "MLP ((50,), relu, sgd, 0.1) (more): 0.9791\n",
      "MLP ((100,), logistic, lbfgs, 0.01) (more): 0.9791\n",
      "MLP ((100,), logistic, adam, 0.01) (more): 0.9791\n",
      "MLP ((100,), relu, lbfgs, 0.1) (more): 0.9791\n",
      "MLP ((150,), logistic, sgd, 0.1) (more): 0.9791\n",
      "MLP ((150,), tanh, lbfgs, 0.1) (more): 0.9791\n",
      "MLP ((150,), relu, lbfgs, 0.1) (more): 0.9791\n",
      "Naive Bayes (more): 0.9783\n",
      "KNN k=3 (more): 0.9783\n",
      "KNN k=5 (more): 0.9783\n",
      "MLP ((50,), logistic, sgd, 0.0001) (more): 0.9783\n",
      "MLP ((50,), logistic, sgd, 0.001) (more): 0.9783\n",
      "MLP ((50,), logistic, sgd, 0.01) (more): 0.9783\n",
      "MLP ((50,), logistic, sgd, 0.1) (more): 0.9783\n",
      "MLP ((100,), logistic, adam, 0.0001) (more): 0.9783\n",
      "MLP ((100,), logistic, adam, 0.001) (more): 0.9783\n",
      "MLP ((100,), tanh, lbfgs, 0.1) (more): 0.9783\n",
      "MLP ((100,), relu, lbfgs, 0.0001) (more): 0.9783\n",
      "MLP ((100,), relu, sgd, 0.0001) (more): 0.9783\n",
      "MLP ((100,), relu, sgd, 0.001) (more): 0.9783\n",
      "MLP ((100,), relu, sgd, 0.01) (more): 0.9783\n",
      "MLP ((100,), relu, sgd, 0.1) (more): 0.9783\n",
      "MLP ((50,), logistic, lbfgs, 0.01) (more): 0.9775\n",
      "MLP ((50,), relu, lbfgs, 0.0001) (more): 0.9775\n",
      "MLP ((50,), relu, lbfgs, 0.01) (more): 0.9775\n",
      "MLP ((50,), relu, lbfgs, 0.1) (more): 0.9775\n",
      "MLP ((100,), logistic, sgd, 0.0001) (more): 0.9775\n",
      "MLP ((100,), logistic, sgd, 0.001) (more): 0.9775\n",
      "Decision tree with pruned branches (more): 0.9766\n",
      "KNN k=1 (more): 0.9766\n",
      "MLP ((50,), tanh, lbfgs, 0.1) (more): 0.9766\n",
      "MLP ((100,), logistic, sgd, 0.01) (more): 0.9766\n",
      "MLP ((100,), logistic, sgd, 0.1) (more): 0.9766\n",
      "MLP ((100,), relu, lbfgs, 0.01) (more): 0.9766\n",
      "MLP ((150,), logistic, lbfgs, 0.0001) (more): 0.9766\n",
      "MLP ((150,), relu, lbfgs, 0.01) (more): 0.9766\n",
      "MLP ((100,), relu, lbfgs, 0.001) (more): 0.9758\n",
      "MLP ((150,), relu, lbfgs, 0.0001) (more): 0.9758\n",
      "MLP ((150,), relu, lbfgs, 0.001) (more): 0.9758\n",
      "MLP ((50,), logistic, lbfgs, 0.0001) (more): 0.9741\n",
      "MLP ((150,), tanh, lbfgs, 0.01) (more): 0.9741\n",
      "MLP ((100,), logistic, lbfgs, 0.001) (more): 0.9733\n",
      "MLP ((100,), tanh, lbfgs, 0.001) (more): 0.9733\n",
      "MLP ((150,), tanh, lbfgs, 0.0001) (more): 0.9733\n",
      "Decision tree without limits (more): 0.9725\n",
      "MLP ((50,), tanh, lbfgs, 0.01) (more): 0.9725\n",
      "MLP ((100,), tanh, lbfgs, 0.0001) (more): 0.9725\n",
      "MLP ((150,), tanh, lbfgs, 0.001) (more): 0.9725\n",
      "MLP ((100,), logistic, lbfgs, 0.0001) (more): 0.9708\n",
      "MLP ((50,), tanh, lbfgs, 0.001) (more): 0.9699\n",
      "MLP ((100,), tanh, lbfgs, 0.01) (more): 0.9699\n",
      "MLP ((50,), tanh, lbfgs, 0.0001) (more): 0.9674\n"
     ]
    }
   ],
   "source": [
    "resultsSorted = dict(sorted(results.items(), key=lambda item: item[1], reverse=True))\n",
    "for key, value in resultsSorted.items():\n",
    "    if \"more\" in key:\n",
    "        print(f\"{key}: {value:.4f}\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The neural network coped with the problem very well. We can see that the selection of parameters did not play a major role here, all combinations obtained satisfactory results. KNN with seven and nine neighbors, also with very good results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Naive Bayes (more): 0.9783\n",
      "Decision tree with pruned branches (more): 0.9766\n",
      "Decision tree without limits (more): 0.9725\n"
     ]
    }
   ],
   "source": [
    "for key, value in resultsSorted.items():\n",
    "    if \"more\" in key:\n",
    "        if \"tree\" in key or \"Bayes\" in key:\n",
    "            print(f\"{key}: {value:.4f}\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Naive Bayes and decision trees also performed well. Interestingly, limiting the tree slightly improved its performance."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "</br>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now let's see less processed data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MLP ((50,), tanh, lbfgs, 0.1) (less): 0.9946\n",
      "MLP ((50,), logistic, lbfgs, 0.1) (less): 0.9938\n",
      "MLP ((50,), tanh, lbfgs, 0.0001) (less): 0.9938\n",
      "MLP ((50,), tanh, lbfgs, 0.001) (less): 0.9938\n",
      "MLP ((50,), tanh, lbfgs, 0.01) (less): 0.9938\n",
      "MLP ((100,), logistic, lbfgs, 0.1) (less): 0.9938\n",
      "MLP ((100,), tanh, lbfgs, 0.1) (less): 0.9938\n",
      "MLP ((150,), logistic, lbfgs, 0.1) (less): 0.9938\n",
      "MLP ((150,), tanh, lbfgs, 0.0001) (less): 0.9938\n",
      "MLP ((150,), tanh, lbfgs, 0.001) (less): 0.9938\n",
      "MLP ((150,), tanh, lbfgs, 0.01) (less): 0.9938\n",
      "MLP ((150,), tanh, lbfgs, 0.1) (less): 0.9938\n",
      "MLP ((150,), tanh, adam, 0.0001) (less): 0.9938\n",
      "MLP ((150,), tanh, adam, 0.001) (less): 0.9938\n",
      "MLP ((150,), tanh, adam, 0.01) (less): 0.9938\n",
      "MLP ((50,), tanh, adam, 0.0001) (less): 0.9930\n",
      "MLP ((50,), tanh, adam, 0.001) (less): 0.9930\n",
      "MLP ((50,), tanh, adam, 0.01) (less): 0.9930\n",
      "MLP ((50,), relu, lbfgs, 0.0001) (less): 0.9930\n",
      "MLP ((50,), relu, lbfgs, 0.01) (less): 0.9930\n",
      "MLP ((50,), relu, lbfgs, 0.1) (less): 0.9930\n",
      "MLP ((50,), relu, adam, 0.0001) (less): 0.9930\n",
      "MLP ((50,), relu, adam, 0.001) (less): 0.9930\n",
      "MLP ((50,), relu, adam, 0.1) (less): 0.9930\n",
      "MLP ((100,), tanh, lbfgs, 0.0001) (less): 0.9930\n",
      "MLP ((100,), tanh, lbfgs, 0.001) (less): 0.9930\n",
      "MLP ((100,), tanh, lbfgs, 0.01) (less): 0.9930\n",
      "MLP ((100,), tanh, adam, 0.0001) (less): 0.9930\n",
      "MLP ((100,), tanh, adam, 0.001) (less): 0.9930\n",
      "MLP ((100,), tanh, adam, 0.01) (less): 0.9930\n",
      "MLP ((100,), relu, adam, 0.0001) (less): 0.9930\n",
      "MLP ((100,), relu, adam, 0.001) (less): 0.9930\n",
      "MLP ((100,), relu, adam, 0.01) (less): 0.9930\n",
      "MLP ((150,), logistic, lbfgs, 0.01) (less): 0.9930\n",
      "MLP ((150,), relu, adam, 0.0001) (less): 0.9930\n",
      "MLP ((150,), relu, adam, 0.001) (less): 0.9930\n",
      "MLP ((150,), relu, adam, 0.01) (less): 0.9930\n",
      "KNN k=1 (less): 0.9923\n",
      "MLP ((50,), logistic, adam, 0.0001) (less): 0.9923\n",
      "MLP ((50,), logistic, adam, 0.001) (less): 0.9923\n",
      "MLP ((50,), tanh, adam, 0.1) (less): 0.9923\n",
      "MLP ((50,), relu, lbfgs, 0.001) (less): 0.9923\n",
      "MLP ((50,), relu, adam, 0.01) (less): 0.9923\n",
      "MLP ((100,), logistic, adam, 0.0001) (less): 0.9923\n",
      "MLP ((100,), logistic, adam, 0.001) (less): 0.9923\n",
      "MLP ((100,), logistic, adam, 0.01) (less): 0.9923\n",
      "MLP ((100,), tanh, adam, 0.1) (less): 0.9923\n",
      "MLP ((100,), relu, lbfgs, 0.1) (less): 0.9923\n",
      "MLP ((100,), relu, adam, 0.1) (less): 0.9923\n",
      "MLP ((150,), logistic, lbfgs, 0.0001) (less): 0.9923\n",
      "MLP ((150,), logistic, adam, 0.0001) (less): 0.9923\n",
      "MLP ((150,), logistic, adam, 0.001) (less): 0.9923\n",
      "MLP ((150,), logistic, adam, 0.01) (less): 0.9923\n",
      "MLP ((150,), tanh, adam, 0.1) (less): 0.9923\n",
      "MLP ((150,), relu, lbfgs, 0.1) (less): 0.9923\n",
      "MLP ((150,), relu, adam, 0.1) (less): 0.9923\n",
      "MLP ((50,), logistic, lbfgs, 0.0001) (less): 0.9915\n",
      "MLP ((50,), logistic, lbfgs, 0.001) (less): 0.9915\n",
      "MLP ((100,), relu, lbfgs, 0.0001) (less): 0.9915\n",
      "MLP ((100,), relu, sgd, 0.0001) (less): 0.9915\n",
      "MLP ((100,), relu, sgd, 0.001) (less): 0.9915\n",
      "MLP ((100,), relu, sgd, 0.01) (less): 0.9915\n",
      "MLP ((100,), relu, sgd, 0.1) (less): 0.9915\n",
      "KNN k=3 (less): 0.9907\n",
      "KNN k=5 (less): 0.9907\n",
      "KNN k=7 (less): 0.9907\n",
      "MLP ((50,), logistic, lbfgs, 0.01) (less): 0.9907\n",
      "MLP ((50,), logistic, adam, 0.01) (less): 0.9907\n",
      "MLP ((100,), relu, lbfgs, 0.001) (less): 0.9907\n",
      "MLP ((100,), relu, lbfgs, 0.01) (less): 0.9907\n",
      "MLP ((150,), logistic, lbfgs, 0.001) (less): 0.9907\n",
      "MLP ((150,), relu, lbfgs, 0.0001) (less): 0.9907\n",
      "MLP ((150,), relu, lbfgs, 0.001) (less): 0.9907\n",
      "MLP ((50,), logistic, adam, 0.1) (less): 0.9899\n",
      "MLP ((100,), logistic, lbfgs, 0.001) (less): 0.9899\n",
      "MLP ((100,), logistic, lbfgs, 0.01) (less): 0.9899\n",
      "MLP ((150,), identity, adam, 0.01) (less): 0.9899\n",
      "MLP ((150,), logistic, adam, 0.1) (less): 0.9899\n",
      "MLP ((150,), relu, lbfgs, 0.01) (less): 0.9899\n",
      "MLP ((150,), relu, sgd, 0.0001) (less): 0.9899\n",
      "MLP ((150,), relu, sgd, 0.001) (less): 0.9899\n",
      "MLP ((150,), relu, sgd, 0.01) (less): 0.9899\n",
      "MLP ((150,), relu, sgd, 0.1) (less): 0.9899\n",
      "MLP ((50,), identity, lbfgs, 0.0001) (less): 0.9892\n",
      "MLP ((50,), identity, lbfgs, 0.001) (less): 0.9892\n",
      "MLP ((50,), identity, lbfgs, 0.01) (less): 0.9892\n",
      "MLP ((50,), identity, lbfgs, 0.1) (less): 0.9892\n",
      "MLP ((50,), relu, sgd, 0.0001) (less): 0.9892\n",
      "MLP ((50,), relu, sgd, 0.001) (less): 0.9892\n",
      "MLP ((50,), relu, sgd, 0.01) (less): 0.9892\n",
      "MLP ((50,), relu, sgd, 0.1) (less): 0.9892\n",
      "MLP ((100,), identity, lbfgs, 0.0001) (less): 0.9892\n",
      "MLP ((100,), identity, lbfgs, 0.001) (less): 0.9892\n",
      "MLP ((100,), identity, lbfgs, 0.01) (less): 0.9892\n",
      "MLP ((100,), identity, lbfgs, 0.1) (less): 0.9892\n",
      "MLP ((100,), logistic, lbfgs, 0.0001) (less): 0.9892\n",
      "MLP ((150,), identity, lbfgs, 0.0001) (less): 0.9892\n",
      "MLP ((150,), identity, lbfgs, 0.001) (less): 0.9892\n",
      "MLP ((150,), identity, lbfgs, 0.01) (less): 0.9892\n",
      "MLP ((150,), identity, lbfgs, 0.1) (less): 0.9892\n",
      "MLP ((150,), identity, adam, 0.0001) (less): 0.9892\n",
      "MLP ((150,), identity, adam, 0.001) (less): 0.9892\n",
      "KNN k=9 (less): 0.9884\n",
      "MLP ((50,), identity, adam, 0.0001) (less): 0.9884\n",
      "MLP ((50,), identity, adam, 0.001) (less): 0.9884\n",
      "MLP ((50,), identity, adam, 0.01) (less): 0.9884\n",
      "MLP ((50,), identity, adam, 0.1) (less): 0.9884\n",
      "MLP ((100,), identity, adam, 0.0001) (less): 0.9884\n",
      "MLP ((100,), identity, adam, 0.001) (less): 0.9884\n",
      "MLP ((100,), identity, adam, 0.01) (less): 0.9884\n",
      "MLP ((100,), identity, adam, 0.1) (less): 0.9884\n",
      "MLP ((100,), logistic, adam, 0.1) (less): 0.9884\n",
      "MLP ((100,), tanh, sgd, 0.0001) (less): 0.9884\n",
      "MLP ((100,), tanh, sgd, 0.001) (less): 0.9884\n",
      "MLP ((100,), tanh, sgd, 0.01) (less): 0.9884\n",
      "MLP ((100,), tanh, sgd, 0.1) (less): 0.9884\n",
      "MLP ((150,), identity, adam, 0.1) (less): 0.9884\n",
      "MLP ((50,), tanh, sgd, 0.0001) (less): 0.9876\n",
      "MLP ((50,), tanh, sgd, 0.001) (less): 0.9876\n",
      "MLP ((50,), tanh, sgd, 0.01) (less): 0.9876\n",
      "MLP ((50,), tanh, sgd, 0.1) (less): 0.9876\n",
      "MLP ((100,), identity, sgd, 0.0001) (less): 0.9876\n",
      "MLP ((100,), identity, sgd, 0.001) (less): 0.9876\n",
      "MLP ((100,), identity, sgd, 0.01) (less): 0.9876\n",
      "MLP ((100,), identity, sgd, 0.1) (less): 0.9876\n",
      "MLP ((50,), identity, sgd, 0.0001) (less): 0.9861\n",
      "MLP ((50,), identity, sgd, 0.001) (less): 0.9861\n",
      "MLP ((50,), identity, sgd, 0.01) (less): 0.9861\n",
      "MLP ((50,), identity, sgd, 0.1) (less): 0.9861\n",
      "MLP ((150,), tanh, sgd, 0.0001) (less): 0.9861\n",
      "MLP ((150,), tanh, sgd, 0.001) (less): 0.9861\n",
      "MLP ((150,), tanh, sgd, 0.01) (less): 0.9861\n",
      "MLP ((150,), tanh, sgd, 0.1) (less): 0.9861\n",
      "MLP ((50,), logistic, sgd, 0.0001) (less): 0.9853\n",
      "MLP ((50,), logistic, sgd, 0.001) (less): 0.9853\n",
      "MLP ((50,), logistic, sgd, 0.01) (less): 0.9853\n",
      "MLP ((50,), logistic, sgd, 0.1) (less): 0.9853\n",
      "MLP ((150,), identity, sgd, 0.0001) (less): 0.9853\n",
      "MLP ((150,), identity, sgd, 0.001) (less): 0.9853\n",
      "MLP ((150,), identity, sgd, 0.01) (less): 0.9853\n",
      "MLP ((150,), identity, sgd, 0.1) (less): 0.9853\n",
      "MLP ((100,), logistic, sgd, 0.0001) (less): 0.9845\n",
      "MLP ((100,), logistic, sgd, 0.001) (less): 0.9845\n",
      "MLP ((100,), logistic, sgd, 0.01) (less): 0.9845\n",
      "MLP ((100,), logistic, sgd, 0.1) (less): 0.9845\n",
      "MLP ((150,), logistic, sgd, 0.0001) (less): 0.9845\n",
      "MLP ((150,), logistic, sgd, 0.001) (less): 0.9845\n",
      "MLP ((150,), logistic, sgd, 0.01) (less): 0.9845\n",
      "MLP ((150,), logistic, sgd, 0.1) (less): 0.9845\n",
      "Decision tree without limits (less): 0.9814\n",
      "Naive Bayes (less): 0.9760\n",
      "Decision tree with pruned branches (less): 0.9598\n"
     ]
    }
   ],
   "source": [
    "for key, value in resultsSorted.items():\n",
    "    if \"less\" in key:\n",
    "        print(f\"{key}: {value:.4f}\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here the situation is completely different. The results are better without z-score and PCA. Data standardization alone was sufficient in this case. Additionally, the best results were achieved by neural networks, although KNN, Trees and Bayes have an accuracy of over 95%."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KNN k=1 (less): 0.9923\n",
      "KNN k=3 (less): 0.9907\n",
      "KNN k=5 (less): 0.9907\n",
      "KNN k=7 (less): 0.9907\n",
      "KNN k=9 (less): 0.9884\n"
     ]
    }
   ],
   "source": [
    "for key, value in resultsSorted.items():\n",
    "    if \"less\" in key:\n",
    "        if \"KNN\" in key:\n",
    "            print(f\"{key}: {value:.4f}\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Error matrices and charts"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's visualize some models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "keysToShow = [\"MLP ((50,), tanh, lbfgs, 0.1) (less)\", \"MLP ((50,), logistic, lbfgs, 0.1) (less)\", \"KNN k=1 (less)\",\n",
    "              \"KNN k=3 (less)\", \"Decision tree without limits (less)\", \"Naive Bayes (less)\"]\n",
    "dictfilt = lambda x, y: dict([ (i,x[i]) for i in x if i in set(y) ])\n",
    "someOfBestResults = dictfilt(results, keysToShow)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decision tree without limits (less) - Accuracy: 0.9814\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Naive Bayes (less) - Accuracy: 0.9760\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KNN k=1 (less) - Accuracy: 0.9923\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KNN k=3 (less) - Accuracy: 0.9907\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MLP ((50,), logistic, lbfgs, 0.1) (less) - Accuracy: 0.9938\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MLP ((50,), tanh, lbfgs, 0.1) (less) - Accuracy: 0.9946\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_confusion_matrix_custom(cm, title='Confusion matrix'):\n",
    "    sns.heatmap(cm, annot=True, cmap='Blues', fmt='d')\n",
    "    plt.ylabel('True')\n",
    "    plt.xlabel('Predicted')\n",
    "    plt.title(title)\n",
    "    plt.show()\n",
    "\n",
    "for clf, name in zip(classifiers, names):\n",
    "    for resultName in someOfBestResults.keys():\n",
    "        if resultName == name:\n",
    "            y_pred = clf.predict(X_test_less)\n",
    "            accuracy = accuracy_score(y_test_less, y_pred)\n",
    "            print(f\"{name} - Accuracy: {accuracy:.4f}\")\n",
    "            \n",
    "            # Visualization of the error matrix\n",
    "            matrix = confusion_matrix(y_test_less, y_pred)\n",
    "            plot_confusion_matrix_custom(matrix, title=name)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The limited memory of the hardware on which the calculations are performed did not allow us to maintain the learning curves. </br>\n",
    "Source: https://github.com/amueller/introduction_to_ml_with_python/issues/67 </br>\n",
    "There will be a learning curve in this work later."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Different classification"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's also see how the models will deal with the classification of the wine quality column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>type</th>\n",
       "      <th>fixed acidity</th>\n",
       "      <th>volatile acidity</th>\n",
       "      <th>citric acid</th>\n",
       "      <th>residual sugar</th>\n",
       "      <th>chlorides</th>\n",
       "      <th>free sulfur dioxide</th>\n",
       "      <th>total sulfur dioxide</th>\n",
       "      <th>density</th>\n",
       "      <th>pH</th>\n",
       "      <th>sulphates</th>\n",
       "      <th>alcohol</th>\n",
       "      <th>quality</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>white</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.27</td>\n",
       "      <td>0.36</td>\n",
       "      <td>20.7</td>\n",
       "      <td>0.045</td>\n",
       "      <td>45.0</td>\n",
       "      <td>170.0</td>\n",
       "      <td>1.0010</td>\n",
       "      <td>3.00</td>\n",
       "      <td>0.45</td>\n",
       "      <td>8.8</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>white</td>\n",
       "      <td>6.3</td>\n",
       "      <td>0.30</td>\n",
       "      <td>0.34</td>\n",
       "      <td>1.6</td>\n",
       "      <td>0.049</td>\n",
       "      <td>14.0</td>\n",
       "      <td>132.0</td>\n",
       "      <td>0.9940</td>\n",
       "      <td>3.30</td>\n",
       "      <td>0.49</td>\n",
       "      <td>9.5</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>white</td>\n",
       "      <td>8.1</td>\n",
       "      <td>0.28</td>\n",
       "      <td>0.40</td>\n",
       "      <td>6.9</td>\n",
       "      <td>0.050</td>\n",
       "      <td>30.0</td>\n",
       "      <td>97.0</td>\n",
       "      <td>0.9951</td>\n",
       "      <td>3.26</td>\n",
       "      <td>0.44</td>\n",
       "      <td>10.1</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>white</td>\n",
       "      <td>7.2</td>\n",
       "      <td>0.23</td>\n",
       "      <td>0.32</td>\n",
       "      <td>8.5</td>\n",
       "      <td>0.058</td>\n",
       "      <td>47.0</td>\n",
       "      <td>186.0</td>\n",
       "      <td>0.9956</td>\n",
       "      <td>3.19</td>\n",
       "      <td>0.40</td>\n",
       "      <td>9.9</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>white</td>\n",
       "      <td>7.2</td>\n",
       "      <td>0.23</td>\n",
       "      <td>0.32</td>\n",
       "      <td>8.5</td>\n",
       "      <td>0.058</td>\n",
       "      <td>47.0</td>\n",
       "      <td>186.0</td>\n",
       "      <td>0.9956</td>\n",
       "      <td>3.19</td>\n",
       "      <td>0.40</td>\n",
       "      <td>9.9</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    type  fixed acidity  volatile acidity  citric acid  residual sugar  \\\n",
       "0  white            7.0              0.27         0.36            20.7   \n",
       "1  white            6.3              0.30         0.34             1.6   \n",
       "2  white            8.1              0.28         0.40             6.9   \n",
       "3  white            7.2              0.23         0.32             8.5   \n",
       "4  white            7.2              0.23         0.32             8.5   \n",
       "\n",
       "   chlorides  free sulfur dioxide  total sulfur dioxide  density    pH  \\\n",
       "0      0.045                 45.0                 170.0   1.0010  3.00   \n",
       "1      0.049                 14.0                 132.0   0.9940  3.30   \n",
       "2      0.050                 30.0                  97.0   0.9951  3.26   \n",
       "3      0.058                 47.0                 186.0   0.9956  3.19   \n",
       "4      0.058                 47.0                 186.0   0.9956  3.19   \n",
       "\n",
       "   sulphates  alcohol  quality  \n",
       "0       0.45      8.8        6  \n",
       "1       0.49      9.5        6  \n",
       "2       0.44     10.1        6  \n",
       "3       0.40      9.9        6  \n",
       "4       0.40      9.9        6  "
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataMore = data.copy()\n",
    "dataMore.fillna(dataMore.median(numeric_only=True), inplace=True)\n",
    "dataMore = dataMore.replace({'type': {'white': 0, 'red': 1}})\n",
    "\n",
    "X_more = dataMore.drop('quality', axis=1)\n",
    "y_more = dataMore['quality']\n",
    "\n",
    "scaler = StandardScaler()\n",
    "X_more = scaler.fit_transform(X_more)\n",
    "\n",
    "X_train_more, X_test_more, y_train_more, y_test_more = train_test_split(X_more, y_more, test_size=0.2, random_state=269402)\n",
    "\n",
    "classifiersQuality = []\n",
    "namesQuality = []"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Decision tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "dt_more_large = DecisionTreeClassifier(random_state=269402)\n",
    "dt_more_large.fit(X_train_more, y_train_more)\n",
    "classifiersQuality.append(dt_more_large)\n",
    "namesQuality.append(f\"Decision tree without limits (more)\");"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Naive Bayes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "nb_more = GaussianNB()\n",
    "nb_more.fit(X_train_more, y_train_more)\n",
    "classifiersQuality.append(nb_more)\n",
    "namesQuality.append(f\"Naive Bayes (more)\");"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K-Nearest Neighbors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "k_values = [1, 3, 5, 7, 9]\n",
    "for k in k_values:\n",
    "    knn_clf = KNeighborsClassifier(n_neighbors=k)\n",
    "    knn_clf.fit(X_train_more, y_train_more)\n",
    "    classifiersQuality.append(knn_clf)\n",
    "    namesQuality.append(f'KNN k={k} (more)')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Neural network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "hidden_layer_sizes = [(100,), (150,)]\n",
    "activations = ['tanh', 'relu']\n",
    "solvers = ['sgd', 'adam']\n",
    "alphas = [0.0001, 0.001, 0.01, 0.1]\n",
    "\n",
    "for hidden_layer_size in hidden_layer_sizes:\n",
    "    for activation in activations:\n",
    "        for solver in solvers:\n",
    "            for alpha in alphas:\n",
    "                mlp = MLPClassifier(hidden_layer_sizes=hidden_layer_size, activation=activation,\n",
    "                                    solver=solver, alpha=alpha, max_iter=15000, random_state=269402)\n",
    "                mlp.fit(X_train_more, y_train_more)\n",
    "                classifiersQuality.append(mlp)\n",
    "                namesQuality.append(f\"MLP ({hidden_layer_size}, {activation}, {solver}, {alpha}) (more)\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "resultsQuality = {}\n",
    "for clf, name in zip(classifiersQuality, namesQuality):\n",
    "    y_pred = clf.predict(X_test_more)\n",
    "    accuracy = accuracy_score(y_test_more, y_pred)\n",
    "    resultsQuality[name] = accuracy;"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KNN k=1 (more): 0.6385\n",
      "Decision tree without limits (more): 0.6108\n",
      "MLP ((150,), relu, adam, 0.0001) (more): 0.5869\n",
      "MLP ((100,), relu, sgd, 0.1) (more): 0.5846\n",
      "MLP ((150,), tanh, adam, 0.01) (more): 0.5815\n",
      "MLP ((150,), relu, sgd, 0.01) (more): 0.5800\n",
      "MLP ((100,), relu, sgd, 0.001) (more): 0.5792\n",
      "MLP ((100,), relu, sgd, 0.01) (more): 0.5777\n",
      "MLP ((150,), tanh, adam, 0.0001) (more): 0.5777\n",
      "MLP ((100,), relu, sgd, 0.0001) (more): 0.5769\n",
      "MLP ((100,), relu, adam, 0.01) (more): 0.5762\n",
      "MLP ((150,), relu, sgd, 0.1) (more): 0.5762\n",
      "MLP ((150,), relu, adam, 0.01) (more): 0.5762\n",
      "MLP ((150,), tanh, sgd, 0.01) (more): 0.5754\n",
      "MLP ((150,), tanh, adam, 0.001) (more): 0.5746\n",
      "MLP ((150,), relu, adam, 0.001) (more): 0.5738\n",
      "MLP ((150,), relu, sgd, 0.001) (more): 0.5700\n",
      "MLP ((150,), tanh, sgd, 0.0001) (more): 0.5677\n",
      "MLP ((150,), tanh, sgd, 0.001) (more): 0.5669\n",
      "MLP ((150,), tanh, sgd, 0.1) (more): 0.5669\n",
      "MLP ((150,), relu, adam, 0.1) (more): 0.5654\n",
      "MLP ((100,), tanh, sgd, 0.01) (more): 0.5638\n",
      "MLP ((100,), tanh, sgd, 0.0001) (more): 0.5623\n",
      "MLP ((100,), tanh, sgd, 0.001) (more): 0.5623\n",
      "MLP ((100,), tanh, sgd, 0.1) (more): 0.5623\n",
      "MLP ((100,), relu, adam, 0.1) (more): 0.5623\n",
      "KNN k=7 (more): 0.5615\n",
      "MLP ((100,), tanh, adam, 0.1) (more): 0.5600\n",
      "MLP ((100,), tanh, adam, 0.0001) (more): 0.5592\n",
      "MLP ((100,), relu, adam, 0.001) (more): 0.5592\n",
      "MLP ((150,), relu, sgd, 0.0001) (more): 0.5592\n",
      "KNN k=5 (more): 0.5585\n",
      "MLP ((100,), tanh, adam, 0.01) (more): 0.5569\n",
      "MLP ((100,), relu, adam, 0.0001) (more): 0.5569\n",
      "MLP ((100,), tanh, adam, 0.001) (more): 0.5562\n",
      "MLP ((150,), tanh, adam, 0.1) (more): 0.5562\n",
      "KNN k=9 (more): 0.5546\n",
      "KNN k=3 (more): 0.5531\n",
      "Naive Bayes (more): 0.4069\n"
     ]
    }
   ],
   "source": [
    "resultsSortedQuality = dict(sorted(resultsQuality.items(), key=lambda item: item[1], reverse=True))\n",
    "for key, value in resultsSortedQuality.items():\n",
    "    print(f\"{key}: {value:.4f}\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When classifying wine quality, our models performed significantly worse than before. The k=1 KNN method is the best, but it could not be fully trusted with an accuracy of 64%."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decision tree without limits (more) - Accuracy: 0.6108\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Naive Bayes (more) - Accuracy: 0.4069\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KNN k=1 (more) - Accuracy: 0.6385\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KNN k=9 (more) - Accuracy: 0.5546\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MLP ((100,), relu, sgd, 0.01) (more) - Accuracy: 0.5777\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MLP ((100,), relu, sgd, 0.1) (more) - Accuracy: 0.5846\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "keysToShow = [\"KNN k=1 (more)\", \"KNN k=9 (more)\", \"MLP ((100,), relu, sgd, 0.01) (more)\",\n",
    "              \"MLP ((100,), relu, sgd, 0.1) (more)\", \"Naive Bayes (more)\", \"Decision tree without limits (more)\"]\n",
    "dictfilt = lambda x, y: dict([ (i,x[i]) for i in x if i in set(y) ])\n",
    "someOfBestResultsQuality = dictfilt(results, keysToShow)\n",
    "\n",
    "for clf, name in zip(classifiersQuality, namesQuality):\n",
    "    for resultName in someOfBestResultsQuality.keys():\n",
    "        if resultName == name:\n",
    "            y_pred = clf.predict(X_test_more)\n",
    "            accuracy = accuracy_score(y_test_more, y_pred)\n",
    "            print(f\"{name} - Accuracy: {accuracy:.4f}\")\n",
    "            \n",
    "            # Wizualizacja macierzy błędów\n",
    "            matrix = confusion_matrix(y_test_more, y_pred)\n",
    "            plot_confusion_matrix_custom(matrix, title=name)\n",
    "            \n",
    "            # Krzywa uczenia się dla sieci neuronowych\n",
    "            if isinstance(clf, MLPClassifier):\n",
    "                plt.plot(clf.loss_curve_)\n",
    "                plt.title(f\"Learning curve - {name}\")\n",
    "                plt.xlabel(\"Iterations\")\n",
    "                plt.ylabel(\"Loss function\")\n",
    "                plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Thank you for taking the time to read this work. </br>\n",
    "#### Report prepared by Kamil Schlagowski"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Bibliography"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Materials used when writing the project:\n",
    "\n",
    "- lectures on computational intelligence\n",
    "\n",
    "\n",
    "- https://www.kaggle.com/datasets/rajyellow46/wine-quality\n",
    "\n",
    "- https://github.com/amueller/introduction_to_ml_with_python/issues/67\n",
    "\n",
    "- https://datascience.stackexchange.com/questions/20199/train-test-split-error-found-input-variables-with-inconsistent-numbers-of-sam\n",
    "\n",
    "- https://jakbadacdane.pl/jak-wyhodowac-drzewo-decyzyjne/\n",
    "\n",
    "- https://scikit-learn.org/stable/modules/preprocessing.html\n",
    "\n",
    "- https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.copy.html\n",
    "- https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.get_dummies.html\n",
    "\n",
    "- https://realpython.com/sort-python-dictionary/\n",
    "- https://realpython.com/python-sort/\n",
    "\n",
    "- https://stackoverflow.com/questions/70897794/finding-the-mean-of-nuisance-columns-in-dataframe-error\n",
    "- https://stackoverflow.com/questions/49871101/how-to-select-range-of-rows-in-pandas#49871201\n",
    "- https://stackoverflow.com/questions/37154329/jupyter-notebook-command-for-hide-the-output-of-a-cell\n",
    "- https://stackoverflow.com/questions/2590503/how-to-find-whether-a-string-is-contained-in-another-string#2590519\n",
    "- https://stackoverflow.com/questions/2590503/how-to-find-whether-a-string-is-contained-in-another-string#2590519\n",
    "- https://stackoverflow.com/questions/3420122/filter-dict-to-contain-only-certain-keys\n",
    "\n",
    "- Github Copilot also helped me:\n",
    "</br> https://github.com/features/copilot\n",
    "- and Chat GPT 4.0:\n",
    "</br> https://chat.openai.com/?model=gpt-4"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.11"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}