Road Accident Risk Prediction Dataset

Table of Contents

• Overview
• Dataset Structure
• Reference
• License

Overview

The Road Accident Risk Prediction dataset is a curated collection of tabular data designed to advance automated prediction of road accident risks. It provides a robust set of samples for regression modeling, enabling researchers and students to explore advanced data processing, feature engineering, and machine learning models for proactive accident prevention and safer road management.

The goal is to predict the likelihood of road accidents (accident_risk) for each road segment.

Property	Value
Total samples	250000
Training samples	200000
Test samples	50000
Features	Tabular data
Target	Continuous value (between 0 and 1)

Dataset Structure

Clone the repository:

git clone [email protected]:AIOZAI/road_accident_risk_data.git

Files

File	Samples	Description
data/train.csv	200000 samples	Training data with features and accident_risk target.
data/test.csv	50000 samples	Test data for submission - accident_risk column hidden.

Data Format

This file contains the features and ground truth labels for the training samples.

Field	Type	Description
id	integer	Unique identifier for each road segment.
road_type	string	Type of road (urban, rural, highway).
num_lanes	integer	Number of lanes on the road.
curvature	float	Degree of road curvature.
speed_limit	integer	Maximum allowed speed limit (km/h).
lighting	string	Lighting condition (daylight, dim, night).
weather	string	Weather condition during observation (clear, rainy, foggy).
road_signs_present	boolean	Indicates whether road signs are present (True or False).
public_road	boolean	Indicates if the road is public (True or False).
time_of_day	string	Time category of the observation (morning, afternoon, evening).
holiday	boolean	Indicates whether it was a public holiday (True or False).
school_season	boolean	Indicates whether the observation occurred during a school season (True or False).
num_reported_accidents	integer	Number of previously reported accidents in the area.
accident_risk	float	Target variable: Continuous value (0–1) representing accident likelihood.

Example: train.csv

id,road_type,num_lanes,curvature,speed_limit,lighting,weather,road_signs_present,public_road,time_of_day,holiday,school_season,num_reported_accidents,accident_risk
0,urban,2,0.06,35,daylight,rainy,False,True,afternoon,False,True,1,0.13
1,urban,4,0.99,35,daylight,clear,True,False,evening,True,True,0,0.35

Reference

For more information about the dataset, please visit the Simulated Roads Accident Data on Kaggle.

License

This dataset is released under the Creative Commons Public Domain Dedication (CC0 1.0).