Death Clock

Disclaimer

Use the pre-trained models instead of building your own model for the hackathon

This repository contains the boilerplate code for the Mortality Dataset. You can refer to the kaggle workspaces to fetch/download your data.

Description of files

FileName	Functionality
App.py	Contains the HTTP server
model-train.ipynb	Notebook for training
model.json	Saved model
model.h5	Saved weights of the previous training
Requirements.txt	Dependencies used

Flow

• Install dependencies

pip install -r requirements.txt

• Test the server

python app.py

• Train the model Open the Jupyter Notebook and tune the model for higher accuracy
• Save the model

model.save()

• Test the model
• Deploy it

Reccomenedations

Once you've completed tuning the model for higher accuracy, you can train the model on:

• Azure ML
• Amazon Sagemaker
• Kaggle Worksapces
• Google's Colab