https://github.com/cinodon/outlier-detection-saas-tfg

Outlier Detection System in SaaS Access Data
https://github.com/cinodon/outlier-detection-saas-tfg

anomaly-detection clustering outlier-detection saas scikit-learn unsupervised-learning

Last synced: about 1 month ago
JSON representation

Outlier Detection System in SaaS Access Data

• Host: GitHub
• URL: https://github.com/cinodon/outlier-detection-saas-tfg
• Owner: cinodon
• Created: 2024-11-04T16:55:12.000Z (about 2 months ago)
• Default Branch: main
• Last Pushed: 2024-11-04T17:49:28.000Z (about 2 months ago)
• Last Synced: 2024-11-04T18:18:34.278Z (about 2 months ago)
• Topics: anomaly-detection, clustering, outlier-detection, saas, scikit-learn, unsupervised-learning
• Language: Python
• Homepage:
• Size: 12.2 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Anomaly Detection in SaaS Access Data

This repository contains the code and resources for my Bachelor's Thesis project on anomaly detection in access data for SaaS (Software as a Service) applications. The goal of the project is to detect unusual patterns in user access logs, which may indicate security issues or abnormal behaviors, by using unsupervised machine learning techniques.

## Project Overview

The system is built around an ETL (Extract, Transform, Load) pipeline in **Python** that processes anonymized access data stored in a **PostgreSQL** database. This data is analyzed through machine learning models to identify potential anomalies. The project utilizes **Docker** for containerization, ensuring consistency across environments, and a **React** front-end interface for visualizing results.

## Features

- **Configurable ETL Pipeline**: The pipeline is managed through a YAML configuration file, allowing customization of database settings, data transformations, and model parameters.

- **Unsupervised Anomaly Detection**: Uses **Isolation Forest**, **DBSCAN**, and **Local Outlier Factor (LOF)** to detect unusual access patterns.

- **Data Visualization**: Generates visual representations of detected anomalies for easier interpretation.

- **Dockerized Deployment**: Encapsulates the application in Docker containers for improved portability and scalability.

## Technologies

- **Python**: Data processing and ETL pipeline development.

- **scikit-learn**: Clustering and outlier detection models.

- **PostgreSQL**: Database management for storing access data.

- **Docker**: Containerizes the application for consistent deployment.

- **React**: Front-end framework for visualizing results.

- **Flask**: Back-end API for data processing and integration with the front-end.

## Installation

1. Clone this repository:

   ```bash

   git clone https://github.com/cinodon/outlier-detection-saas-tfg.git

   cd tfg-iam-risk-pusture-poc

2. Set up and run Docker containers

   ```bash

   docker compose up

3. Access the app

   : http://localhost:3000