S&P500 Stock Market Index Price Prediction

Project Overview

In this project, we'll predict the price of the S&P500 stock market index.

Project Steps

• Download data using the yfinance package
• Create an initial machine learning model and estimate accuracy
• Build a backtesting engine to more accurately measure accuracy
• Improve the accuracy of the model

Features

• Automatic Data Fetching: Retrieves S&P 500 historical data from 1980 to the current date using the yfinance library.
• Feature Engineering: Dynamically computes technical indicators like:
  • Close_Ratio_2, Close_Ratio_5, Close_Ratio_60, Close_Ratio_250, and Close_Ratio_1000.
• Model Training and Evaluation:
  • Implements Time-Series Cross-Validation for robust evaluation.
  • Tracks key metrics including Precision, Recall, F1-Score, and ROC-AUC.
• Interactive Visualizations:
  • Time-Series Chart: Shows historical S&P 500 trends.
  • Feature Importance: Highlights the most impactful features using an interactive bar chart.
• Buy/Sell Recommendations:
  • Provides actionable insights with confidence levels for the latest available data.

File overview:

• market_prediction.ipynb - a Jupyter notebook that contains all of the code.

Live Application

Check out the live Streamlit app:
Stock Prediction App

Technical Overview

Libraries Used

• Streamlit: For building the interactive web app.
• yfinance: To fetch historical S&P 500 data.
• scikit-learn: For machine learning and evaluation metrics.
• pandas: For data manipulation and feature engineering.
• plotly: For interactive visualizations.

Model Overview

The Random Forest Classifier predicts whether the stock price will rise or fall based on historical trends. The app provides recommendations as:

• Buy: When the model predicts a price increase with a confidence level above the user-defined threshold.
• Sell: When the model predicts a price decrease.

---

App Preview

Key Sections

1. Historical Performance:
   • Displays the historical trends of the S&P 500.
   • Helps users understand past market behavior.
2. Model Metrics:
   • Evaluates model performance using robust metrics like Precision, Recall, F1-Score, and ROC-AUC.
3. Buy/Sell Insights:
   • Provides clear, actionable recommendations for decision-making.

---

Contributing

Contributions are welcome! Feel free to open an issue or submit a pull request.

Installation

To follow this project, please install the following locally:

• JupyterLab
• Python 3.8+
• Python packages:
  • pandas
  • yfinance
  • scikit-learn