This project predicts the posibilities of a telecoms company churning. It makes use of
the following deep learning aspects:
    SelectKBest (feature selection)
    Funtional API (model creation)
    Activation functions (mainy Relu)
    GridSearch for Hyperparameter tuning

User inputs 7 different inputs based on their profiles on the flask app platform
The input is then converted into a datafram and parsed to the trained model for prediction
The returned probability is then converted into a 0 or 1
A zero represents a low probability of churning whie a 1 represents otherwise
The app displays the prediction as well as the prediction confidence

Here is the link to the short demonstration: https://drive.google.com/file/d/1_WGcA0OFC2qhqJC2q9cZc4abGBZoGAcw/view?usp=sharing

It is also found in the DemoVideo.Zip file.