【摘要】 随着计算机技术的进步，机器学习在多个领域的应用日益广泛。在地震工程领域，如何有效利用机器学习以解决实际问题，已成为地震工程专家关注的焦点。本文立足于18届世界地震工程大会中的会议论文，以近年来国内外相关研究文献作为补充，对基于机器学习的结构地震响应预测和损伤评估的相关研究进行了总结和评述。首先，从算法的角度分别梳理了机器学习和深度学习在结构地震响应预测方面的研究现状，介绍了现有研究中的常用算法及其适用性；其次，按照数据类型将数据分为时序数据和图像数据，针对每一类数据，分别总结评述了其基于机器学习的结构地震损伤评估的研究现状，包括数据来源、研究流程以及优缺点；最后，针对目前存在的数据质量不高或分布不均衡、参数选择困难和模型泛化性能较差等问题，讨论了未来的研究方向，旨在推动机器学习在地震工程领域的深入应用和进一步发展。更多还原

【Abstract】 With the advancement of computer technology, machine learning(ML) is becoming increasingly widespread in many fields. In the field of earthquake engineering, how to effectively use ML to solve practical problems has become the focus of attention of earthquake engineering experts. Based on the conference papers in the 18th World Conference on Earthquake Engineering(18WCEE), this paper summarized and commented on the related research of structural seismic response prediction and damage assessment based on ML, with the relevant research literature at home and abroad in recent years as a supplement. Firstly, from the perspective of algorithms, the research status of ML and deep learning(DL) in structural seismic response prediction was reviewed, and the commonly used algorithms and their applicability in existing research were introduced. Secondly, according to the data type, the data was divided into time series data and image data. For each type of data, the research status of structural seismic damage assessment based on ML was summarized and reviewed, including data source, research process, advantages and disadvantages. Finally, in view of the existing problems such as low data quality or uneven distribution, difficult parameter selection, and poor model generalization performance, the future research directions were discussed, aiming to promote the in-depth application and further development of ML in the field of earthquake engineering.更多还原