报告题目:机器学习方法在气象领域的应用探索研究
报告人:于廷照 高级工程师(中国气象局公共气象服务中心)
报告时间:2023年6月30日10:30
报告地点:秀山校区逸夫楼202会议室
报告对象:全校感兴趣的师生
主办单位:计算机科学与技术学院
报告人简介:于廷照,中国气象局公共气象服务中心高级工程师,中国气象局重点创新团队“能源气象团队”成员,博士毕业于中国科学院自动化研究所模式识别国家重点实验室,长期从事“气象+机器学习”模型算法研究。2022年入选中国气象局高层次科技创新人才计划“青年气象英才”,现为中国气象服务协会人工智能技术委员会管理专家组成员。主持、参加国家级、省部级项目10余项,发表SCI论文20余篇,获授权气象相关国家发明专利8项,软件著作权9项。
报告摘要:In recent years, the utilization of machine learning methods in the field of meteorology has gained widespread attention as a research hotspot. This report delves into the similarities between machine learning problems and meteorological problems, and highlights our endeavors to leverage these similarities in areas like meteorological downscaling, image-based weather recognition, meteorological nowcasting, and short-term meteorological forecast. Our results showcase that the application of machine learning methods can yield impressive prediction and recognition results in meteorology, thereby providing significant support and guidance to the fields of meteorological forecasting and disaster prevention and mitigation. Finally, we discuss the future research, practical directions, and prospects for leveraging machine learning methods in meteorology.