주행의도 판단을 위해 Sequential Backward Selection을 이용한 최적의 특 징값 선정 방법

Abstract

Many researchers have attempted to model driver behavior for predicting the trajectory of surrounding vehicles. It is difficult to predict, even a few seconds in advance, since it depends on each driver’s intention and driving habits. In an effort to cope with these obstacles, various approaches using NGSIM datasets and machine learning algorithm have been proposed. However, many literature use different feature and algorithm to predict driver’s intention. In this paper, we present a novel features to be used in machine learning algorithms to estimate lane change intention.