Portfolio

Investing in the stock market can be challenging for individual investors who lack access to sophisticated analytical tools and expertise. PredictWise addresses this problem by combining historical price data, technical indicators, and machine learning to forecast future stock movements. I developed this comprehensive system that fetches real-time market data, calculates 30+ financial indicators (moving averages, RSI, MACD, etc.), and trains multiple ML models (RandomForest, XGBoost) to predict price changes. Drawing inspiration from financial analysis techniques and modern ML practices found on platforms like Kaggle and research papers, I integrated these concepts into an intuitive Streamlit dashboard where users can visualize trends, compare stocks, and access predictions. The system employs a modular design with separate components for data collection, feature engineering, model training, and visualization. You can find a detailed Medium article exploring the technical architecture and lessons I learned here!

Github