Inspiration
Seeing how we can make an impact by predicting renewable energy investments for each state appealed to us.
What it does
We created a linear regression model using sklearn in Python that takes in 18 features according to the MSN's and outputs the predicted amount of investment given the input.
How we built it
We trained the linear regression model on 250 data points (50 states * 5 years), using the MSN input for each state/year combination, and tested it on the 2020 test data set.
Challenges we ran into
We tried a neural network model, but the predictions were not accurate as there were not enough data points. We also tried logistic regression and LASSO models, but they did not produce accurate models.
Accomplishments that we're proud of
We are proud that, with a simple linear regression model, we were able to produce interpretable results.
What we learned
We learned the general data science analysis pipeline and how to clean data in order to analyze it.
What's next for Data Dudes Chevron Project
More fun projects!
Log in or sign up for Devpost to join the conversation.