Inspiration
Our interest in cars, specifically in highly sought after startups that have failed.
What it does
Builds a prediction model of a dataset we created to help predict a model's profitability in the modern market
How we built it
We built a random forest model to help classify each model as either profitable or non-profitable. After that we dropped the lowest contributing feature to help tune the model
Challenges we ran into
There were no available data sets for us to utilize specifically on electric vehicle, so we had to acquire our own. On top of that, we weren't able to use a web scraper to collect a large sample size. Another huge challenge was changing our model type towards the end of the project. We had planned for a linear regression, but unfortunately our data did not fit, meaning we had to end up with a classification model last second.
Accomplishments that we're proud of
We helped define our own metric for affordability, which was pretty interesting. It's also a huge accomplishment for else turning in a project as this is our first hackathon and data science project (Completely new to the subject)
What we learned
Plan better before moving on to the next step, having to back trace after logging in hours of continuous work is super frustrating
What's next for ElectricPredictionModel
Improve on some of our faults, it's no secret we have a small sample size and even looked into synthetic data generation, but ultimately could never figure it out. We need to find a better way to standardize the affordability metric and add on a couple more features to increase the reliability of our target variables. Hopefully in the future, we can use it to product how a startup will fare in the market!
Log in or sign up for Devpost to join the conversation.