Hey there! Welcome to my (absolutely not created last night) school project where I tackled the age-old problem of... length of stays. I mean, who wants to overstay their welcome? Certainly not me. And certainly not the 30,000 test cases we had.
- A couple of
.csv
files with mystery data. (Well, it's not THAT mysterious. One is for training, the other for testing. But let's keep the air of mystery going.) - A Python script that looks suspiciously like I actually did something!
- My results, which prove that I didn't just watch Netflix while munching popcorn during the semester.
- Loaded some data.
- Preprocessed like a boss.
- Trained using the mighty Random Forest Regressor because...trees, right?
- Validated my model and saved it so it doesn't run away.
pandas
: For handling my data and ensuring I don't get bamboozled.joblib
: Because who wants to train a model every single time?!sklearn
: My partner-in-crime for everything ML.
This project contains over 70,000 training cases. That's almost the population of a small city. They now know more about length of stays than most of us!
This wonderful masterpiece is brought to you by Birkeland. Shout out to caffeine for making this possible!