Date: 20 May 2021

Speaker: André Rios, Eren Aktas and Dr Bojan Bozic

Title: Unsupervised Learning for Anomaly Detection

Abstract:

The purpose of this project is to investigate unsupervised learning techniques for anomaly detection – particularly for utilizing self organizing maps and auto-encoders. The ideal outcome of the project is a comprehensive state of the art review – combined with a sample, early phase s/w output that illustrates a state of the art development in the area related to application that uses structured data or time series data.

Online Recording:  https://youtu.be/hVzp43PUacI