Improving accuracy in road accidents prediction by comparing support vector machine with Lasso regression.

The primary purpose of this investigation is to compare and contrast the performance of two machine learning algorithms, namely the Lasso regression Algorithm and the Support Vector Machine Algorithm, in the context of accident prediction. Over the course of 20 iterations, utilising N=10 samples, the Support Vector Machine approach, and Lasso regression, the accuracy of accident predicting was tested. A substantial gap (84.28 percent) separates the performance of the Support Vector Machine method from that of the Lasso regression methodology (94.56 percent). A significance level of 0.001 has been assigned to the new SVM classifier (p0.05 independent sample test). Comparisons of data showed that the Support Vector Machine Novel classifier was more accurate than the Lasso regression in predicting the occurrence of traffic accidents. [ABSTRACT FROM AUTHOR]