Energy Consumption Anomaly Detection

This repository contains the work done by group 17 in the subject IT2901 in 2022. The project aims to use machine learning to detect anomalies in energy consumption. The project is done for Trondheim municipality and the data used are from their buildings.

Anomaly Detection Pipeline

Here we deploy our Energy Temperature Time (ETT) model. The pipeline runs in Orchest and integrates with Clarify. The anomalies found can be viewed in Clarify. The pipeline is located in this repo: https://github.com/dagokl/trdk02_energy_consumption_anomalies_pipeline

Notebooks

In addition to the pipeline, we have created some notebook while experimenting with different models. These notebooks are located in the models folder.