ML for everybody

This project aids in extending the boundaries of machine learning not just to the ones with the ability to code. This project is built in Pyqt and provides a simple interface for performing data preprocessing tasks and machine learning tasks on any dataset.

How to run

After cloning the repository.

cd ml_for_everybody
python codes/uicode.py

Screenshots

User Interface. User can use the buttons to load a dataset and to begin the data preprocessing tasks.

Once data is loaded, user can see the dataset and the datatypes. User can choose to convert categorical values/drop columns or even scale numerical values.

Once the user is happy with the dataset, they can choose to train a machine learning model on it. We have given option between Linear regression, logitic regression and neural networks. This is the screen for neural networks.

Here the user can set the parameters for the model. From learning rate to the strucute of the model can also be decided by the user.

Additionally we also have provided the user with a correlation heatmap for reducing collinearity and reduce overfitting.