Lokari - Web server log anomaly detector

NOTE: Maintenance ongoing! Creating a new repo to work with!

Machine learning based monitoring application for anomalies in web server logs.

Machine learning project by Tuomo Kuure (tqre) and Joni Hakala (Dunttus).
Supervising teacher: Tero Karvinen - Haaga-Helia University of Applied Sciences

Original project blog: https://ailogs.design.blog/
This repository is a cleaned version of the original: https://github.com/Dunttus/AI-Project

Directories:

/.idea - PyCharm configuration files
/autoencoder - Autoencoding ML layer construction
/data_processing - Data processing modules
/datasets - Datasets used in training
/docker - Project runtime container

Files:

config.py - Main configuration file
main.py - Log monitor
train.py - Detector training

Requirements

• Docker
• nvidia-container-toolkit
• PyCharm Professional for development