This data set is collected from Addis Ababa Sub-city police departments for master's research work. The data set has been prepared from manual records of road traffic accidents of the year 2017-20. It has 32 features and 12316 instances of the accident.
How to approach a Machine Learning project:-
Exploratory Data Analysis
Preprocessing
Feature Transformation
Baseline Modeling
Hyperparameter Tuning
Explainable AI
Deployment a Model ( Streamlit App deployment on Heroku) : https://road-traffic-severity-predict.herokuapp.com
Traffic accident prediction is a crucial and challenging issue in the domain of intelligent traffic safety management system;
It is of great significance for analyzing the future development trend of traffic accidents and implementing proactive prevention measures under existing road traffic conditions.
This is a multi-class classification problem where we are predicting the severity of accident :
- Slight Injury
- Fatal Injury
- Serious Injury
based on the other 31 features.