In this competition you must predict whether a mushroom is edible or poisonous using the tabular data you are provided with.
My approaches have been:
-FastAI TabularPandas to prep the tabular data and then fed into a sklearn RandomForestClassifier, I have gotten about a 0.52 on my best submission which is pretty bad since it is not any better than random guessing.
-From-scratch neural network which accuracy ranges between 0.45 and 0.52
-XGBoost Classifier has gotten me a 0.97 accuracy submission, took more care of the data before feeding it to the model (see latest commit for mushroomxgbm.ipynb)