Research and Application of Intersection Similarity Algorithm Based on KNN Classification Model

The study on the characteristics of traffic intersections can solve the traffic jam, route design and other important traffic problems. This paper classifies the intersection features based on machine learning algorithm. The KNN classification model is trained to classify the traffic import data collected, and the similarity of the intersection is calculated and matched according to the feature extraction and weighting technology. Thus, a new algorithm of intersection similarity matching is established, and it is studied and applied. The algorithm can match similar intersections, and provide model and theoretical support for traffic department and related technicians.