AWS-ML-Hackathon

This repository contains machine learning model approaches created by Syed Rizvi for AWS-NVIDIA Environmental Hackathon in August, 2020.

Autoencoder approach placed 3rd overall, awarded $3000 prize.

Approaches

• Autoencoder time-series reconstruction for anomaly detection (Placed 3rd overall in competition)
• Isolation Forest Algorithm

Autoencoder Reconstruction Sample

Requirements

• Pytorch
• Pandas
• Numpy
• Sklearn
• Matplotlib

Demo video

Link to demo video: https://youtu.be/mYcjEGL58d8