This project analysed the Kaggle playground challenge Kobe Bryant's shot selection. The data contained the location and circumstances of every field goal attempted by Kobe Bryant during his 20-year career. The task was to predict whether the basket went in (shot_made_flag). Kaggle had removed 5000 of the shot_made_flags (represented as missing values in the csv file). These were the test set shots for which one must submit a prediction.
Please find "main.pdf" for the summary report of this analysis.
.
+-- _main.Rmd
+-- _data <- The dataset provided by Kaggle for the challenge
+-- _code <- Jupyter notebooks and Python/R scripts
+-- _submission <- submission files to the competition
+-- main.pdf <- The main PDF report produced by R Markdown
Various models were run for this project:
- XGBOOST in Python
- XGBOOST in R
- TensorFLOW DNN in Python
- sklearn (random forest, logistic regression, linear discriminant analysis, K-NN, decision tree, naive bayes, extra tree, adaboost, and gbm) in Python
The following tools/platforms were used for this analysis:
- Digital Ocean
- Kaggle's kernel
- Google Cloud Platform's docker
- Python and R (with Jupyter Notebooks, RStudio, and RMarkdown)
- x86_64-apple-darwin13.4.0 (64-bit)
I would like to credit the following Exploratory Data Analysis notebooks upon which these analyses were built: