Pre-Collision Warning and Recommendation System for Assistant Driver using Least Square Support Vector Machine and Fuzzy Logic
Alifia Puspaningrum, Adi Suheryadi, A Sumarudin · 2019
1.25 million people lose their lives due to road accidents every year. To deal with this problem, some innovative transportation are proposed to avoid road accidents, one of them is giving alert and recommendation for each vehicle which is well known as Pre-Collision Warning. According to many researches that have been conducted before, the performance of pre-collision warning can be improved by considering not only speed, but also the traffic sign and the distance to other objects. Furthermore, after predicting the danger, the system needs to give the speed recommendation and its action. This research conducts Least Square - Support Vector Machine (LS-SVM) to predict the danger and give the speed recommendation using Fuzzy Logic. The experiment shows that the best performance is achieved by LS-SVM according to the value of 98.71% for accuracy, 98.60% for precision, and 99.18% for recall.