AI-Driven Trading System using Ridge Regression for 3-day recursive price forecasting. It features VROC (Volume Rate of Change) to filter market noise and identifying false signals. Built with Python & Streamlit, it delivers real-time interactive charts, currency conversion, and ~57% directional accuracy for smarter, data-driven decisions.
A Real-Time Financial Dashboard that uses Recursive Machine Learning (Ridge Regression) to predict stock prices for the next 3 days.
- About the Project
- Key Features
- Tech Stack
- System Architecture
- Installation & Setup
- How It Works
- Future Scope
- Contact
Retail traders often struggle with emotional decision-making and reliance on lagging indicators (like simple Moving Averages) that only reflect past market behavior. Existing tools are either too simplistic or expensive "black-box" systems.
This project solves that gap.
It is a Hybrid Analytical System that combines:
- Technical Analysis: Uses proven indicators like RSI, EMA, and VROC (Volume Rate of Change).
- Machine Learning: Implements Ridge Regression with L2 Regularization to handle market noise.
- Recursive Forecasting: Instead of just predicting tomorrow, it predicts T+1, T+2, and T+3 days ahead to show the trend direction.
- 🔮 Recursive Multi-Step Forecasting: Predicts stock prices for the next 3 days (72 hours) using a recursive loop strategy.
- 📊 Interactive Real-Time Charts: Features professional-grade, zoomable candlestick charts powered by
lightweight-charts. - 🧠 Noise Reduction AI: Uses Ridge Regression to filter out multicollinearity and prevent overfitting on volatile data.
- 📉 Smart VROC Filtering: Validates price breakouts by analyzing Volume Rate of Change—ignoring "fake" pumps.
- 🌍 Universal Asset Support: Works with Stocks (Reliance, Apple), Crypto (Bitcoin, Ethereum), and Forex via Yahoo Finance.
- 💱 Currency Converter: Automatically converts USD assets to INR (₹) for Indian traders.
| Component | Technology | Description |
|---|---|---|
| Language |  |
Core Logic & Scripting |
| Frontend |  |
Web Dashboard UI |
| ML Core |  |
Ridge Regression Model |
| Data |  |
Data Manipulation & Time-Series |
| API | yfinance | Live OHLCV Market Data |
The system follows a 4-Layer Architecture:
- Data Layer: Fetches live data from Yahoo Finance API.
- Processing Layer: Cleans data, fixes Timezones (UTC -> IST), and calculates Technical Indicators.
- Intelligence Layer: Runs the Ridge Regression model to generate the 3-Day Forecast.
- Presentation Layer: Displays the Dashboard and Signals via Streamlit.
(You can upload your architecture diagram image here)
Follow these steps to run the project locally.
- Python 3.8 or higher installed.
-
Clone the Repository
git clone [https://github.com/your-username/AI-Trading-System.git](https://github.com/your-username/AI-Trading-System.git) cd AI-Trading-System -
Install Dependencies
pip install -r requirements.txt
-
Run the Application
streamlit run app.py
-
Access the Dashboard Open your browser and go to
http://localhost:8501.
- Select Asset: Enter a ticker symbol (e.g.,
BTC-USDorRELIANCE.NS). - Data Fetch: The system pulls the last 60 days of OHLCV data.
- Feature Engineering: It calculates RSI, EMA, and VROC.
- Training: The Ridge Model trains on this data instantly.
- Forecasting:
- Predicts Price for Day 1.
- Appends Day 1 prediction to data -> Predicts Day 2.
- Appends Day 2 prediction to data -> Predicts Day 3.
- Visualization: Plots the historical data and the future 3-day trend line.
- Sentiment Analysis: Integrating News API to analyze market sentiment.
- Deep Learning: Testing LSTM models for long-term trend analysis.
- Live Alerts: Sending Email/WhatsApp alerts when a Buy signal is detected.
Name: [Bhukya Ram Ganesh] Email: [bhukyaramganesh7897@gmail.com] Project Link: https://github.com/your-username/AI-Trading-System