TradingIntelligence is a quick and simple repository used for testing different machine learning models on stock market data with the purpose of predicting short-term stock price changes.
The dataset used in the project is the popular S&P 500 stock data from Kaggle. To limit computational requirements, the following 10 stocks were chosen for the project: AAPL, AMZN, BRK-B, DIS, GOOG, JPM, META, MSFT, NVDA, and TSLA. The dataset contains stock data from 2013 to 2018.
The following models were tested on the dataset:
- Linear Regression
- Support Vector Machine (SVM)
- Long Short-Term Memory (LSTM)
The range of models provides a sufficient variety of complexity.
The models.ipynb notebook contains the code used for executing the models. Each model is implemented in the src directory.
To use the Jupyter notebook and test the models, the following procedure can be followed:
- Install the required packages (it is recommended to use the Conda environment in the
investing.yamlfile). - Run the Jupyter notebook cell-by-cell.
The notebook contains diagrams and success metrics (i.e., accuracy, precision, recall, and F1 score) for each model.