BachelorTree is a silly project. My girlfriend really seems to like watching ABC's show, "The Bachelor". Hence, I sometimes watch it since I enjoy her company. The show itself is a fairly silly, but I'm always curious as to whether or not there's any way to answer a few questions:
- Why do some girls stay?
- Why do some girls go?
- Who will go next? (Boo! Hiss!! Predictive analysis!!!)
http://blog.thehumangeo.com/2015/02/20/machine-learning-about-love-who-will-leave-the-bachelor-next/
- We write some sort of data scraper.
- We use decision trees* to analyze the data we've captured
- We watch the show next week and realize how wrong we are.
I've written a data scraper that hits the ABC website and pulls some data down. I've jammed all that into a JSON file that has today's date on it in /feature_data/.
- You can give me realistic advice on better methods.
- You can add or enhance the data about the contestants.
- You can write some code to do something you think is necessary.
- Twitter: jimmysthoughts
*Yes, I know, decision trees on a dataset of 24 points with many features is a horrible, horrible idea.