DIV

A two-stage volatility forecasting system for BANKNIFTY options using GARCH(1,1) + XGBoost. The pipeline ingests raw NSE options data, computes implied volatility, builds a daily feature table, fits a rolling GARCH model, and then trains an XGBoost model to correct GARCH's systematic blind spots using options market signals.

This project, BadhKavach, is a flood prediction model built using various geospatial and machine learning techniques. The model uses terrain data, height maps, and segmentation maps to simulate flood scenarios and predict flood-prone areas.