Website link : dolab.virgyferdian.my.id
This project is an end-to-end Machine Learning web application designed to classify next-day stock price movements as Uptrend or Downtrend. Developed as my undergraduate thesis (skripsi) at Politeknik Negeri Malang, the system moves beyond simple price regression by focusing on actionable binary classification for the Indonesian Stock Exchange (IDX).
Full-Stack Implementation
- Backend: Flask (Python) micro-framework managing the API and model inference.
- Data Pipeline: Integrated Yahoo Finance API for real-time data ingestion of IDX stocks.
- Frontend: A responsive dashboard built with TailwindCSS and Alpine.js, featuring persistent prediction history stored via browser local storage.
- Deployment Tools: TensorFlow for the ML engine and WandB for experiment tracking and artifact management.
Screenshots:
