Market Manipulation Detection System

Project Overview

This capstone project develops a sophisticated machine learning system to detect market manipulation patterns in high-frequency trading environments. The project integrates real-time market data, order book analysis, and sentiment analysis to identify potentially manipulative trading behaviors across multiple securities.

!Market Analysis Visualization

Key Results & Findings

Data Quality Metrics

• Accuracy: 99.97% price accuracy validated against Yahoo Finance
• Dataset Scale: 55,508 minute-level observations across 35 days
• Securities: TSLA (41.1%), AAPL (34.5%), MSFT (24.3%)

Technical Achievements

• Multi-API Integration: Alpaca Markets, Interactive Brokers, Alpha Vantage
• Database Architecture: PostgreSQL 3NF design with three core tables
• Real-time Pipeline: Automated data collection with error handling
• Feature Engineering: Manipulation detection algorithms

Technologies: Python, PostgreSQL, Alpaca API, IBKR API, Alpha Vantage, VADER EOF < /dev/null