This project compares, evaluates, and predicts ELAN stock price using Monte Carlo Simulation, FB Prophet, SMA, k-Nearest Neighbors, Arima, LSTM, and EMA.
Content incudes:
Include but not limited to the following:
This project leverages python 3.7
-
Alpaca API - Collect live stock information
-
Monte Carlo Forecast Tools - Visualize all possiblities
-
Prophet - Time series forcasting
-
Google Colab - Jupyter Notebook function in browser
-
Pandas - Data analysis and manipulation
-
Scikit-learn - Supplies machine learning libraries for Python
-
Keras - Provides deep learning API for Python interface
-
Yahoo Finance - Aquire stock data from Yahoo! Finance
-
Tensorflow - Supplies machine learning libraries for Python
- Comparing ELAN and LLY performance
- Predict ELAN cumulative returns
- Explore Machine Learning models:
- Simple Moving Average (SMA)
- k-Nearest Neighbor
- Auto ARIMA
- Long Short Term Memory (LSTM)
- Perform future price predictions with the best model (LSTM):
- Predict future 300 days price using historical daily closing price
- Predict future 56 weeks performance using historical weekly average closing price
LSTM machine learning model performs the best among all explored models, thus, future predictions will be performed on LSTM model.
Yanjun Lin Andrie | email: yanjun.lin.andrie@gmail.com |
UC Berkeley Extension