The Stock-Predictor-V4 Project has undoubtedly reached its full potential, and at this stage, third-party pull requests represent the sole avenue for introducing new features and functionalities. Unfortunately, due to my limited expertise in prediction logic, I had to remove it entirely from the update cycle. Embracing the contributions of skilled developers through pull requests is now the most promising way to breathe fresh life into the project and ensure its continued growth and relevance.
- Stock-Predictor-V4
- 1a. Installation
- 1b. Updating the Repository
- 2. Data Preparation
- 3. Training the LSTM RL Model
- 4. Evaluating the Model
- 5. Fine Tuning the LSTM RL Model
-
- Utilizing the Model for Stock Market Prediction (Is in Constuction, dont use it!)
- 7. Comparing the predicted values with the actual values after the 30-day period.
Recommended PC Specifications for Stock-Predictor-V4:
- GPU VRAM: 4 GB
- RAM: 6 GB (when equipped with a GPU)
- RAM: 10 GB (without a GPU)
- Supported Operating Systems: Linux, Mac, or Windows
If your PC meets at least 70% of these recommendations, it should perform well during the training process. If your specs fall below this threshold, consider running the program on Colab (highly recommended) or on your local machine (May take long).
git clone https://github.com/lex-hue/Stock-Predictor-V4.git
cd Stock-Predictor-V4
python SPV4.py --install
python SPV4.py --update
python SPV4.py --install
*You must retrain your Models so that the new Code Supports it
To prepare the data for stock prediction, you can follow either of the following options:
- Run the
SPV4.pyscript by executingpython SPV4.py --prepare_dataand type in the Ticker Symbol youd like to Preprocess (For Example BTC-USB, AAPL or so).
Upon completion of these steps, your Preprocessed stock data will contain indicators that enhance the reliability of predictions.
To train the LSTM RL Model with the data.csv file that was generated by the 2. Step, execute the following:
python SPV4.py --train
After running this command, the LSTM RL model will start training with the data in the data.csv file.
To evaluate the trained model, execute the following:
python SPV4.py --eval
After running this command, the root mean squared error (RMSE), the MAPE and Total Rewards will be plotted.
To fine tune the model, execute the following:
python SPV4.py --fine_tune
Before starting the fine-tuning process for the LSTM RL model, the script will now ask you to specify your desired reward threshold. It is recommended to set the threshold at 0.9.
It is recommended to maintain hydration by drinking water while waiting for the fine-tuning process to complete, as this may take a while. Once the fine-tuning is done, attempt to run the script with a single loop. If the script doesn't loop once initially, keep rerunning it until it completes at least one loop. If the script consistently produces the same result without any improvement, please interrupt it by pressing CTRL + C, answer "yes" to the prompt, and rerun the script.
After the fine-tuning process is fully complete, it is highly recommended to re-evaluate the model's performance.
If you've reached the end of the 30-day predicted period and you're curious about the accuracy of the model, then the script can help you. To compare the data, you need to update the CSV file you selected in Step 2.
After that, run the following command in your terminal:
python SPV4.py --compare
This will compare the predicted data with the actual data. You may regret some of the decisions you made and wonder why you didn't trust the model (just kidding, don't take it too seriously!).
