Abstract
Vehicle crashes occur because of numerous factors. It leads to loss of lives and permanent incapacity. Due to vehicle crashes, the budgetary expenses for both individuals as well as for the nation are influenced from the vehicle crashes. According to Road accidents statistics a total of 4,64,910 road accidents were reported in India, claiming 1,47,913 lives and causing injuries to 4,70,975 persons, which translates into 405 deaths and 1,290 injuries each day from 1,274 accidents.
In recent years many researches have been carried out and many models like Sequential models, Complex Linear Models, Time Sequence Models etc. has been created for analysis and characterizing purpose. Several Data Mining Techniques are also available, and it deals with two important steps i.e. Prediction of severity and identifying important factors for the accident. Several Machine Learning algorithms are available among those Artificial Neural Networks (ANN), Linear Regression (LR), Random Forest, Decision Trees, SVM are widely used. These are the algorithms which are widely used for predicting severity of the accidents. Tools available are Weka, Tangara, Knime, Orange, R, NLTK. The widely used tools are R and Weka.
This review paper discusses about the algorithm that shows better performance in the earlier researches done, and the widely used tools. The models are implemented based on the performance of the model of the previous research in order to overcome the shortfalls of the earlier work and the data set is taken from Kaggle