ya332 / Churn-Rate

Churn rate calculation via Artificial Neural Networks

Run from Scratch

• First, download Python from here: https://www.python.org/downloads/windows/
• Then, download get-pip.py from https://bootstrap.pypa.io/
  $ cd /path/to/downloadsfolder
$ python get-pip.py
• Install Git from here: https://gitforwindows.org/
• Install requirements
  $ pip install -U -r requirements.txt
• At this point, you should be able to run the neural network
  $ python ann.py

Included Files

• ann.py
• Churn_Modelling.csv
• model.json
• model.h5
• requirements.txt
• README.md
• group1-shard1of1.bin (Keras model is converted to Tensorflow layers so that it could be imported and run on a web browser.)

To convert the model to a tensorflowjs model (For web development only)

$ pip install tensorflowjs
$ tensorflowjs_converter --input_format keras path/to/my_model.h5 path/to/tfjs_target_dir \