Machine Learning in Action

快速开始 - Getting Started • 内容 - Table of Contents • 关于 - About • 鸣谢 - Acknowledgment • FAQ •

Made by ximing Xing • 🌌 https://ximingxing.github.io/

Machine-Learning-in-Action是基于Peter Harrington的<<机器学习实战>>这本书, 将书中的机器学习算法和案例以scikit-learn的代码组织形式呈现.

除了算法机器学习算法之外，更重要的是算法的使用场景，这个代码仓库中也提供机器学习实战案例。包括：百万英文新闻文本分类实战等.

Machine-Learning-in-Action is based on Peter Harrington`s <> , The machine learning algorithms and cases in the book are presented in the form of scikit-learn code organization.

📋 快速开始 - Getting Started

1. Using Pycharm with conda plugin IDE makes getting started easier.

   • check out from version control.

   • chose Git.

2. python setup.py --develop

内容 - Table of Contents

^{▴ Back to top}

├── LICENSE
├── README.md
├── data
│   └── 20news-bydate_py3.pkz
├── examples
│   └── 20newsgroup_in_action
├── mlic
│   ├── cluster
│   ├── linear_model
│   ├── metrics
│   ├── naive_bayes
│   ├── neighbors
│   ├── neural_network
│   ├── svm
│   ├── tree
│   └── utils
├── requestments.txt
├── setup.py
└── tests
    ├── Bayes
    ├── KNN
    └── Linear

关于 - About

完整的数据挖掘过程 :

1. 网络爬虫 Network Crawler

   • 按分类爬取环球网英文本新闻(处理静态网站与需要js渲染的内容)
   • Scrapy-Splash based Crawler crawls information from globaltimes.cn
   • CNN Crawler
   • BBC Crawler

2. 文本分类实战

   • DataLoader : 20newsgroup
   • Data preprocessing
     • Data Cleaning: regular expression
     • Data Cleaning: stop words
     • Normalization: lemmatization
   • Extracting Features : Word Dict and TF-IDF
   • Model: bayes and svm
   • Evaluation

   SVM Pipline

   

   Classification Report

   
   

   Confusion Matrix

   

你无需担心example/与tests/中案例所使用的数据集，因为数据集都是自动下载的.

如果有问题也希望你指出.

鸣谢 - Acknowledgment

^{▴ Back to top}

• 代码组织参考scikit-learn中的组织结构

💬 FAQ

^{▴ Back to top}