Tesla Stock Predictor
- Category: Machine Learning
- Technologies: Python, Scikit-learn, Pandas, Matplotlib, Numpy
- Skills: Pair Programming, Project Management
- Repository: Github Link
- Report: Read Here
In this project, a daily sentiment score was calculated based on the number of likes and retweets on Elon Musk's tweets for each day dating back to 2017. These figures were collected via Twitter's API. Tesla's daily market value data was also collected from Yahoo Finance dating back to 2017.
A Lasso Regression model and a kNN Regression model were trained using the data, with daily tweet sentiment scores as input values and Tesla's daily market values as target values. The training data was randomly split into a test set and a training set using a 5-fold holdout method. Finally, a regressor that predicts the median value was used as a baseline predictor.