Coursera_Capstone Project

This is a repository for my IBM Data Science Professional Certificate - Applied Data Science Capstone projects.

Accident Severity Prediction Using Data Mining Techniques

Introduction / Business Problem

Traffic accidents are undesirable and unforeseen events that have negative consequences like injury, damage and even death which are mainly caused by human, vehicle, road and/or environmental factors. The problem of traffic accident in the world affects people, properties and the nation as a whole, the impact of which leads to death or serious injury, destruction of property and economic loss. A successful solution would be to predict the severity of accidents and also to determine the major factors that leads to traffic accidents in order to readdress those impacts and prevent traffic accidents.

Objectives

1. Explore the variables that impact accident severity.
2. Predicting accident severity using classification models (Random Forest, Naive Bayes, Random Forest).

Data Understanding

Data for this study was collected by the Seattle Police Department and recorded by Traffic Records Group which consist of collisions (traffic accident) data from 2004 to present. You can find the Dataset by Clicking Here. The dataset contains 194,673 rows and 37 columns (features). The predictor or target variable will be ‘SEVERITYCODE’ because it is used to measure the severity of an accident. The feature description of the dataset can be found by Clicking Here.