Tesla Stock Price Prediction through MFCC Feature Extraction and LSTM Modeling

Features

• Converts Tesla stock price data into audio signals using MFCC feature extraction.
• Trains LSTM and SVM models to predict stock price movements based on extracted features.
• Provides a comprehensive analysis of model performance using various metrics.
• Includes both code and datasets required to replicate the study.

Installation

1. Clone the repository:

   git clone https://github.com/jeanqazxcv/master-s-thesis.git

2. Navigate to the project directory:

   cd master-s-thesis

3. Install dependencies:

   pip install -r requirements.txt
   

Usage

1. Run Jupyter Notebooks:

• Open the Jupyter notebooks in the Code/ directory to start running the experiments. You can use the following command to start Jupyter Notebook:

  jupyter notebook
  

• The following .ipynb files are available:
  • mfcc_5daysfeature_tsla_LSTM.ipynb: For training and evaluating the LSTM model.
  • mfcc_5daysfeature_tsla_svm.ipynb: For training and evaluating the SVM model.
  • mfcc_5daysfeature_tsla_svm_pca.ipynb: For training and evaluating the SVM model with PCA applied to the features.

2. Load the Datasets:

• The datasets are located in the Datasets/ directory and include .wav files representing Tesla's stock prices.

3. Train the Models:

• Open the respective Jupyter notebooks to preprocess the data, extract MFCC features, and train the models.
• Follow the instructions in each notebook for specific steps.

Credits

This project is part of a Master's thesis by Jean Hong, focusing on innovative methods for stock price prediction using audio signal processing techniques.