test.csv contains all the information about Kobe Our goal is to use the features and predict his shot type.
The mail idea of our project is using feature shot distance to split thte data into two sets
For example, We split the data with shotdistance greater than 10 and build a model to train these two datasets
We had tried many classifier such as Knn,Logistic Regression , Decision Tree, Xgboost... as we find out that Randomforest has the best performance on both accuracy and running time
We eventually use the model , and tuning parameters based on this model