This project aims to predict whether the closing price of Tesla stock will be higher the next day using historical stock data. It's a binary classification problem where machine learning models help predict a buy signal based on past stock price data.
https://tesla-stock-price-prediction-project.streamlit.app/
The dataset includes Tesla's daily stock prices from January 2010 to December 2017. Key features include:
- Open: Opening price
- High: Highest price
- Low: Lowest price
- Close: Closing price
- Volume: Total shares traded
- Open-Close Spread: Price difference between open and close
- High-Low Spread: Volatility indicator
- Quarter-End Flag: Indicator for quarter-end days
- Logistic Regression
- Support Vector Machines (SVM)
- XGBoost
Model performance is evaluated using ROC-AUC, trained on 90% of the data with 10% reserved for validation.