Energy Consumption Fraud Detection

Inspiration

Fraud detection is a very important domain that can take advantage of machine learning.

What it does

Fraud detection machine learning model that predicts fraudulent electricity or gas consumption.

How we built it

• utilized multiple approaches for improving the evaluation metrics such as feature engineering, undersampling, oversampling, hp tuning on multiple ml models, anomaly detection models

What's next for Energy Consumption Fraud Detection

• Build a fraud detection system for a user interface
• More research in the topic and find potential areas of improvement

Built With

• matplotlib
• numpy
• optuna
• pandas
• python
• sklearn
• sqlite