PREDICTING NBA PLAYER SALARIES BASED ON THEIR PERFORMANCE ON THE COURT USING RANDOM FOREST
The National Basketball Association (NBA) is one of the most popular sports leagues in the world. Due to its rapid development and increasing income, the value of player contracts has increased substantially. There have been various discussions as to whether a player's contract is in accordance...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/55232 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | The National Basketball Association (NBA) is one of the most popular sports leagues in the world. Due to its rapid development and increasing income, the value of player contracts has increased substantially. There have been various discussions as to whether a player's contract is in accordance with his performance on the pitch or not. This research is made with the aim of participating in the discussion using a mathematical method, namely machine learning, or more specifically Random Forest. The purpose of this study is to determine the overpaid and underpaid players and determine which variables most influence the salaries of NBA players. Data is taken from basketball-reference.com and hoopsype.com. Three Random Forest models were created with the metric being OOB RMSE. The best OOB RMSE result was 0,04385142. From this model, a table showing the player's salary category is created. The most important variables to determine NBA player salaries according to the models are Age, Minutes Played per game, Points per game, and Field Goals made per game. |
|---|