A traffic pattern behavior recognition method and corresponding recognition model construction method

Abstract

The invention provides a method for building a traffic pattern behavior recognition model, comprising: 1) carrying out multi-level classification to specific traffic pattern behaviors, and establishing a corresponding classification tree; 2) for the classification problem of each parent node in the classification tree, based on The random forest model selects the optimal feature set corresponding to the test sample of the classification problem according to the recognition accuracy; 3) For the classification problem of each parent node in the classification tree, based on the optimal feature set corresponding to the parent node obtained in step 2), Feature set, select an optimal classification model as the sub-classification model corresponding to the parent node. The invention also provides a corresponding traffic pattern behavior recognition method. The present invention can more accurately distinguish specific traffic mode behaviors, and has less computational complexity.

Description

technical field [0001] The present invention relates to the technical fields of pervasive computing, mobile Internet and urban planning, and more specifically, the present invention relates to a traffic pattern behavior recognition method and a corresponding recognition model construction method. Background technique [0002] Using sensory data to identify users' daily behavior is an important research problem in the field of ubiquitous computing. As a subcategory of many daily behaviors, traffic mode behavior (walking, biking, bus ride, etc.) contains a lot of information related to the user's movement trajectory and life rules. Accurately and effectively capturing users' daily traffic patterns and behaviors, and then analyzing user behavior habits and daily life trajectories, plays an important role in many fields such as smart mobile services, health monitoring, and urban planning. For example, using traffic pattern behavior recognition technology, users' daily trajector...

AI Technical Summary

Problems solved by technology

However, there are many types of traffic mode behaviors, the similarity between different traffic mode behaviors is high, and the internal differences of the same traffic mode behavior are large. These factors lead to the unsatisfactory recognition accuracy of the current traffic recognition schemes. The more difficult it is to meet the requirements of practical applications

Images