Hitters Salary Analysis - README

Overview

This project analyzes the relationship between baseball player statistics and their salaries using the Hitters dataset, which contains information on Major League Baseball players from the 1986 and 1987 seasons. The analysis focuses on predicting salary levels (Low, Medium, High) based on performance metrics, career statistics, and other relevant variables.

Key Features

• Data Preprocessing: Handles missing values, detects outliers, and standardizes numeric predictors.
• Exploratory Data Analysis: Includes visualizations of distributions, correlations, and relationships between variables.
• Multinomial Logistic Regression: Predicts salary levels using performance metrics as predictors.
• Model Evaluation: Uses cross-validation and performance metrics (accuracy, precision, recall, F1-score) to assess model quality.
• Multicollinearity Handling: Addresses correlated predictors using Principal Component Analysis (PCA) and stepwise selection.
• Comprehensive Reporting: Detailed R Markdown document with visualizations, interpretations, and conclusions.

Files

• ProjectWork_GroupE.Rmd: R Markdown file containing the full analysis, including code, visualizations, and explanations.
• ProjectWork_GroupE.html: HTML output of the R Markdown report.
• ProjectWork_GroupE.pdf: PDF version of the HTML report.
• Hitters.csv: Dataset used for the analysis.

How to Use

1. Ensure you have R and RStudio installed.
2. Install the required R packages listed in the setup chunk of the R Markdown file.
3. Place Hitters.csv in your working directory.
4. Open ProjectWork_GroupE.Rmd in RStudio and knit the document to generate the HTML report.

Dependencies

R packages:

• dplyr
• ggplot2
• kableExtra
• corrplot
• reshape2
• psych
• DT
• gridExtra
• nnet
• caret
• car
• MASS
• tidyverse

Authors

• Muhammet Emin Albayram (303991)
• Ege John Isik (302991)