Skip to content

Latest commit

 

History

History
83 lines (63 loc) · 3.99 KB

README.md

File metadata and controls

83 lines (63 loc) · 3.99 KB
Raw

🤖 Predict4me

A web Artificial Intelligence model for everyone - just upload your dataset and let it predict for you

ℹ️ ATTENTION: The fork on @gcbsdev is meant to be read-only and its purpose is for streaming the app to Streamlit Community Cloud. When submitting a Pull Request or opening and Issue, please always refer to the repo in @gabrielgcbs

❓ What is Predict4me?

Predict4me is a web Machine Learning app built with Python and Streamlit

It implements a simple user-friendly interface that let's you upload your data and click one button to train a model and predict the data.

Note that the purporse of Predict4me is to help those who are not experienced with Machine Learning predict their data without needing to code a model, thus it is not suited for more complex predictions or model engineering.

Currently, the app covers only the two Supervised Learning problems, which are:

  • Classification
  • Regression

The models implemented are:

  • Random Forest (for classification tasks)
  • Histogram-based Gradient Boosting Regression Tree (for regression tasks)

ℹ️ For more information about how these models work, please refer to scikit-learn official documentation: RandomForest, HistGradientBoostingRegressor

🏁 Quick Start

🏎️ How to run the app on your browser?

  • Visit predict4me app
  • Select the type of task you wish: Classification or Regression (not sure which one to choose? Have a look at this tutorial)
  • Upload your data:
    • The requirements to get the model running on your data are:
      • File format: csv
      • The data needs to have at least one feature column
      • The data needs to have only one target column (this may change in the future)
      • Nominal categorical features must be specified, if there are any
      • Numerical categorical features do not need to be specified
      • It is advised to clean the data before jumping on the model
    • You must submit one file for training the model (step 1), and one file containg the actual data you want to predict (step 2)
      • For example, if you want to predict the house pricing on a city, provide a dataset with information about houses in the area and their prices on step 1, and a dataset with the information about the house you want to predict on step 2
  • If there are any nominal features (i.e. strings), provide their names on step 3
  • Click on RUN MODEL
  • Save the results as a csv file

💻 How to install and run the app locally?

Install Python >= 3.9

Clone the repository:

$ git clone https://github.com/gabrielgcbs/predict-for-me

ℹ️ Recommended: Create and activate a virtual environment

On Windows:

C:\> python -m venv .venv
C:\> .venv\Scripts\activate.bat

On Linux:

$ python -m venv .venv
$ source .venv/bin/activate

Install the dependencies:

$ pip install requirements.txt

Run the app:

$ streamlit run src/main.py

❤️ Guidelines for contributing

Contributions are welcomed. For that, please follow this simple guidelines:

  • Check if there is already an issue opened
  • If not, create an issue to describe your request and specify a label for it, e.g bug, feature or documentation
  • If you want to make some changes on the code, create a pull request and link it to an existing issue:
    • Documentation is important, so for each significant changes made on the code, please provide a documentation about it, both in-code and on the pull request itself