Traffic Accident Risk Prediction Using Machine Learning

IEEE BASE PAPER ABSTRACT:

The occurrence of road accidents continues to be one of the prominent causes of deaths, disabilities and hospitalization in the country. This makes traffic accident risk prediction important in order to minimise it and save lives. Several kinds of models have been proposed to achieve the same ranging from old statistical models to the new models motivated by the advent of machine learning. This paper presents a comparative study of a variety of these models in an effort to analyse and deduce a beneficial approach to traffic accident risk prediction. Since the drivers are the ones in control on the road the study aims to provide traffic accident risk prediction to the drivers by analysing the factors they would know of beforehand like vehicle type, age sex, time of the day and weather etc. Optimal Classification Trees is a model that would provide such results that make intuitive sense to the driver along with the use of Random Forest and Logistic Regression. Furthermore, the geo-location data analysis using K-means clustering algorithm can provide information regarding places that are more prone to accidents. Through the analysis of previously known factors using these algorithms the drivers can be equipped with traffic accident risk predictions that would help them make informed decisions to minimise the same.

PROJECT OUTPUT VIDEO:

ALGORITHM / MODEL USED:

Random Forest Classifier.

OUR PROPOSED ABSTRACT:

Due to a growth in the number of vehicles, the road has become complex in design and management sectors. Since they significantly affect people’s safety, health, and well-being, traffic accidents are a major cause for concern on a global scale. The World Health Organization (WHO) estimates that 1.35 million individuals lose their lives in traffic accidents every year. As a result, they provide a significant area of study for the application of cutting-edge methods and algorithms for their analysis and prediction.

While many traffic accidents are external, some are internal to the driver. For instance, unfavourable weather conditions like rain, clouds, and fog reduce visibility and make driving on such roads challenging and sometimes dangerous. The existing system prediction model considered only different possible causative factors but they have not considered the geo-location data which has latitude and longitude information.

This project aims to develop a machine learning model for predicting traffic accident risk using a Random Forest Classifier algorithm. The dataset used in this study consists of historical traffic accident data, which includes information about the geo-location, time, location, weather conditions, and other relevant factors associated with each accident.

The dataset is referred from the Kaggle which hosts a data set by the government of the United Kingdom. The Random Forest Classifier algorithm is trained on the dataset, and its ability to predict the likelihood of an accident occurring is evaluated. The performance of the model is assessed using various evaluation metrics such as accuracy, precision, recall, and F1-score.

The results show that the Random Forest Classifier algorithm can effectively predict the risk of traffic accidents with a high degree of accuracy. The study concludes that the Random Forest Classifier algorithm is a promising approach for predicting traffic accident risk, and it could be used to develop effective strategies for preventing accidents and reducing their impact.

SYSTEM REQUIREMENTS:

HARDWARE REQUIREMENTS:

System : Pentium i3 Processor.
Hard Disk : 500 GB.
Monitor : 15’’ LED
Input Devices : Keyboard, Mouse
Ram : 4 GB

SOFTWARE REQUIREMENTS:

Operating System : Windows 10 / 11.
Coding Language : Python 3.8.
Web Framework : Flask.
Frontend : HTML, CSS, JavaScript.

REFERENCE:

Kakoli Banerjee, Vikram Bali, Aanchal Sharma, Deepti Aggarwal, Aakash Yadav, Anirudh Shukla, Prateek Srivastav, “Traffic Accident Risk Prediction Using Machine Learning”, 2022 International Mobile and Embedded Technology Conference (MECON), IEEE Conference, 2022.

Python IEEE Projects

Traffic Accident Risk Prediction Using Machine Learning