Data Calibration Based on Multisensor Using Classification Analysis: A Random Forests Approach

Abstract

This paper analyzes the problem of meaningless outliers in traffic detective data sets and researches characteristics about the data of monophyletic detector and multisensor detector based on real-time data on highway. Based on analysis of the current random forests algorithm, which is a learning algorithm of high accuracy and fast speed, new optimum random forests about filtrating outlier in the sample are proposed, which employ bagging strategy combined with boosting strategy. Random forests of different number of trees are applied to analyze status classification of meaningless outliers in traffic detective data sets, respectively, based on traffic flow, spot mean speed, and roadway occupancy rate of traffic parameters. The results show that optimum model of random forest is more accurate to filtrate meaningless outliers in traffic detective data collected from road intersections. With filtrated data for processing, transportation information system can decrease the influence of error data to improve highway traffic information services.

Multi-source Traffic Data Calibration with Optimized Adaboost

“…The researchers proposed the method of ear-neighbor clustering and genetic algorithm [3], and some scholars put forward the method of evaluating the abrupt data in abnormal traffic flow [4] and the detector data evaluation based on rough set fuzzy recognition method [5] recently. The author of this paper also starts with the method of random forest [6] to give out the corresponding method to verify the outlier data. This paper proposes a new method based on AdaBoost optimization, which is an iterative classification algorithm with high accuracy and fast computation speed, to filter the outliers of the outliers, which is meaningful for traffic data detection.…”