ReadMe file:

Shapley Analysis with Random Forest on Traffic Accident Severity

Introduction

This project presents a comprehensive analysis of traffic accident severity using the Shapley value analysis in combination with a Random Forest model. Our dataset encompasses various factors contributing to accidents, enabling us to predict and understand the severity of these incidents better.

Dataset

The dataset, Accdataset_hk_PS_BAEL_Combined.csv, includes detailed records of traffic accidents, featuring variables such as accident location, type, causes, road features, and weather conditions. Our target variable is Accident_Severity_C, categorized into several severity levels.

Analysis Overview

We employ Random Forest classifiers and regressors to model the accident severity based on a range of predictors excluding identifiers and remarks. The Shapley value analysis further allows us to interpret the model by quantifying each feature's contribution to the prediction.

Key Steps:

• Data preprocessing: Fill missing values and select relevant features.
• Model training: Utilize Random Forest for both classification and regression tasks.
• Hyperparameter tuning: Conduct GridSearchCV to find optimal model parameters.
• Shapley value computation: Utilize shap.TreeExplainer to calculate Shap values for model interpretation.
• Visualization: Generate plots like waterfall and summary plots to visualize the impact of features.

Installation

Ensure you have the following Python libraries installed:

• pandas
• numpy
• sklearn
• shap

You can install these packages using pip:

pip install pandas numpy sklearn shap