Stock Forecast App

An end-to-end forecasting tool: pull historical market data, engineer features, fit time-series models, and serve predictions through an interactive UI.

The pipeline

• Data: automated retrieval and cleaning of historical price series.
• Modeling: classical and ML-based forecasting, evaluated against naive baselines (the honest test most forecast demos skip).
• Interface: an interactive app where a user picks a ticker and horizon and sees the forecast with its uncertainty.

Built as a learning project in applied ML, so the emphasis is on a clean pipeline and honest evaluation, not on beating the market.