AI-Driven Algorithmic Trading Model

This repository contains a machine learning-based algorithmic trading model that predicts stock market actions (Buy or Sell) based on historical stock data. The project includes data preprocessing, feature engineering, model training, and a graphical user interface (GUI) for real-time predictions.

---

Features

• Data Preprocessing: Scales and cleans historical stock data for analysis.
• Feature Engineering: Adds moving averages (MA_10, MA_50) for stock trend analysis.
• Machine Learning Model: Utilizes a Random Forest classifier to predict stock actions.
• Model Persistence: Saves the trained model and scaler for future predictions.
• GUI: User-friendly interface for testing the model on new stock data.

---

Demo

GUI

The project provides a GUI to input stock features and receive predictions:

• Input stock data (e.g., Open, High, Low, Close, Volume, MA_10, MA_50).
• Get a "Buy" or "Sell" prediction instantly.

Screenshot: Sample GUI Screenshot:

---

Project Structure

├── data/
│ └── trading_data.csv # Example dataset
├── gui/
│ └── trading_gui.py # GUI implementation
├── models/
│ ├── model.pkl # Trained machine learning model
│ ├── scaler.pkl # Preprocessing scaler
├── scripts/
│ ├── train_model.py # Model training script
│ ├── prediction.py # Script for making predictions
├── requirements.txt # Python dependencies
├── README.md # Project documentation

---

Getting Started

Follow these steps to set up and run the project locally.

1. Clone the Repository

Clone the repository to your local machine:

git clone https://github.com/adarshpheonix2810/ai-algorithmic-trading.git
cd ai-algorithmic-trading
pip install -r requirements.txt
python scripts/train_model.py

2. Install Dependencies

Install the required Python libraries:

pip install -r requirements.txt

3. Prepare the Dataset

Place your stock market dataset (trading_data.csv) in the data/ directory.
The dataset must include the following columns:

• Open, High, Low, Close, Volume, MA_10, MA_50, Target

If you're using the sample dataset provided, skip this step.

4. Train the Model

Run the training script to train the Random Forest model:

python scripts/train_model.py

This will:

• Train the machine learning model.
• Save the model as model.pkl in the models/ directory.
• Save the scaler as scaler.pkl in the models/ directory.

5. Test the GUI

Launch the GUI to make predictions:

python gui/trading_gui.py

6. Make Predictions via Script (Optional)

You can also make predictions programmatically using the prediction.py script:

from scripts.prediction import predict_stock_action

# Example stock data input
input_data = [[150.0, 155.0, 148.0, 152.0, 1000000, 151.0, 150.5]]
prediction = predict_stock_action(input_data)
print("Prediction:", "Buy" if prediction == 1 else "Sell")

---

Requirements

This project requires Python 3.8+ and the following libraries:

• numpy
• pandas
• scikit-learn
• joblib
• tkinter
• matplotlib
• yfinance (optional, for fetching real-time stock data)

All dependencies are listed in requirements.txt. Install them using:

pip install -r requirements.txt

---

Dataset Overview

Columns Description:

• Open: The opening price of the stock.
• High: The highest price of the stock during the trading day.
• Low: The lowest price of the stock during the trading day.
• Close: The closing price of the stock.
• Volume: The total number of shares traded during the day.
• MA_10: 10-day moving average of the closing price.
• MA_50: 50-day moving average of the closing price.
• Target: A binary label where 1 represents "Buy" and 0 represents "Sell".

---

How the Model Works

Dataset:

• Features: Open, High, Low, Close, Volume, MA_10, MA_50
• Target: 1 for "Buy" and 0 for "Sell"

Preprocessing:

• Data is cleaned and scaled using StandardScaler.

Training:

• A Random Forest classifier is trained on the preprocessed data.

Prediction:

• The trained model predicts "Buy" (1) or "Sell" (0) based on input features.

GUI:

• Allows users to test predictions interactively by inputting stock data.

---

Project Highlights

1. Real-World Applicability: Simulates an algorithmic trading decision-making system.
2. Interactive GUI: Makes the model accessible for non-technical users.
3. Extensible: Easily integrate real-time stock data from APIs like Yahoo Finance or Alpha Vantage.

---

License

This project is licensed under the MIT License. See the LICENSE file for more details.