This project applies Dynamic Mode Decomposition (DMD) to financial time series for stock price prediction and trend analysis. By modeling the stock market as a dynamical system, DMD decomposes complex price movements into growth and decay modes, revealing underlying temporal patterns that drive market behavior.

Using NSE500 stock data, the model achieved 1.94% MAPE, showcasing high short-term accuracy. A Streamlit-based GUI was built for visualizing DMD modes, eigenvalue spectra, and automated stock ranking based on dominant growing modes, enabling interpretable and data-driven trading insights.