This is a piece of software that aims to predict stock prices using ML models and a popular deep learning frameowrk called PyTorch.
Here are some things we aim to continue adding to this project:
- Dynamically changing the sequence length (also known as a look back window)
- Dynamically calculating how much we need to scale the data in order to get the best results
- Using more data such as indicators and incorporating that into our model
- Creating a front end web app to make it easier to interact with the model and see the predictions
- Potentially creating a simple NLP model
Prerequesites (everything listed below will already be included in Google Colab):
- Jupyter Notebooks
- Python
- Open the
model.ipynb
file to find the main ML model and click "Run All". Then you will see the predicted price (by default the close price) at the bottom. - You can also take a look into
prophet.ipynb
which uses a different library to predict stocks and shows the data in a different way.
If you'd like to chat with others working on this project or would like help then feel free to visit our discord server.
Thank you so much for your interest in Quantum Trader! If you'd like to add a feature or report a bug then feel free to open a pull request describing your problem or what you'd like to do.