Anomaly detection in user behavior of websites using Hierarchical Temporal Memories : Using Machine Learning to detect unusual behavior from users of a web service to quickly detect possible security hazards.

Abstract: This Master's Thesis focuses on the recent Cortical Learn-ing Algorithm (CLA), designed for temporal anomaly detection. It is here applied to the problem of anomaly detec-tion in user behavior of web services, which is getting moreand more important in a network security context. CLA is here compared to more traditional state-of-the-art algorithms of anomaly detection: Hidden Markov Models (HMMs) and t-stide (an N-gram-based anomaly detector), which are among the few algorithms compatible withthe online processing constraint of this problem. It is observed that on the synthetic dataset used forthis comparison, CLA performs signicantly better thanthe other two algorithms in terms of precision of the detection. The two other algorithms don't seem to be able tohandle this task at all. It appears that this anomaly de-tection problem (outlier detection in short sequences overa large alphabet) is considerably different from what hasbeen extensively studied up to now.