anomaly-detection-algorithm

This repository is showcasing our Anomaly Detection System, developed as our final project in the software engineering course, utilizing basic statistical techniques like mean, variance, and covariance to detects anomalies

An end-to-end Machine Learning pipeline for Credit Card Fraud Detection. Features Supervised Learning (XGBoost optimized with CUDA GPU & Random Forest) alongside Unsupervised Anomaly Detection techniques.

Comprehensive anomaly detection algorithm designed to analyze and identify anomalies in computer internal features.

Code and data for MMAING-Arbo, a Mixed Model of Artificial Intelligence and Next-Generation that performs anomaly detection in time series data.

Log analysis project aimed at finding and predicting anomalies in logs

Algorithms for the R environment that are able to detect high-density anomalies. Such anomalies are deviant cases positioned in the most normal regions of the data space.

SCR# 3056.0: WGAN-based Digital Twins for Anomaly Detection

Hybrid statistical and LSTM-based models for anomaly detection in time-dependent dynamical systems.

Advanced anomaly detection system using graph neural networks and time series analysis to identify fraudulent transactions, money laundering patterns, and market manipulation in real-time financial data streams.

An end-to-end AI-driven sales forecasting and demand intelligence dashboard built with Python, Streamlit, and Prophet.

EPITA group project, anomaly detection

A Stock Anomaly detection is a project for learning the detection of abnormal instances, called anomalies (or outliers) in the stock market. You’ll design a warning system that will alert regulators of stock price manipulation. This project has applications in data cleaning and detecting fraud.

Project Anomaly Detection (AD), MSc Optional, Year 1, Semester 1, Faculty of Mathematics and Computer Science, University of Bucharest

Product Inspection with FOMO AD (Visual Anomaly Detection) by Edge Impulse on Sony Spresense camera and LCD 1602

Anomaly detection with SECODA for the R environment. SECODA is a general-purpose unsupervised non-parametric anomaly detection algorithm for datasets containing numerical and/or categorical attributes.

A ML project that identifies smoke anomalies in real-time data and sends alert through buzzers. Designed for early detection in safety monitoring systems.

Physics-grounded fault detection and diagnosis for a collaborative robot: unsupervised PCA detector (AUROC 0.944) and Random Forest diagnosis on voraus-AD telemetry.

Advanced anomaly detection using topological data analysis and manifold learning.

An open-source, ultra-fast transaction reconciliation & anomaly detection engine for Banking, Mobile Money, and Telco APIs.

Sovereign, climate-aware anomaly detection for critical infrastructure. One file. Zero dependencies. Zero data leaves your machine.