Utility Evaluation and Optimization of Machine Learning in Intelligent Transportation Systems

With the urbanization process and the continuous growth of traffic demand, intelligent transportation system plays an important role in solving traffic problems. As the core technology of artificial intelligence, machine learning can improve the operation efficiency of traffic system, reduce traffic congestion and reduce the incidence of traffic accidents. This paper summarizes the application and optimization of machine learning in intelligent transportation system, and puts forward the methods and steps of utility evaluation and optimization effect evaluation. By analyzing research conclusions and lessons learned, the paper emphasizes the importance of data quality and scale, algorithm selection, model interpretability and robustness, and interdisciplinary collaboration and comprehensive optimization. However, this paper also points out some shortcomings, such as data quality, algorithm interpretability and privacy protection, which need to be further studied and discussed.