题目:COVID-19 Pandemic Forecasting in Europe: A Transfer Learning with Weights Model Averaging Approach
主讲人:中国科研DB视讯(中国)数学与系统科学DB视讯(中国) 张新雨研究员
主持人:西南财经大学统计学院 常晋源教授
时间:2020年4月17日(星期五)10:00-11:00
直播平台及会议ID:腾讯会议,426 462 927
报告摘要:
We forecast the daily confirmed new cases of the COVID-19 pandemic in major European countries using Transfer Learning (TL) and an model averaging (MA) approach. The TL-MA outperforms other single transfer learning models and provides accurate and stable forecasts. The forecasting is especially accurate for those countries which have implemented inadequate containment measures. For those countries which have adopted effective containment measures at earlier stages, all forecasting models lose certain a level of accuracy.
主讲人简介:
张新雨,中国科研DB视讯(中国)数学与系统科学DB视讯(中国)/预测中心研究员。在中科院系统所取得博士学位学位,曾在TAMU做博士后研究。主要从事计量经济学和统计学的理论和应用研究工作,具体研究方向包括模型平均、模型选择和组合预测等。担任期刊《Journal of Systems Science and Complexity》领域主编、期刊《SADM》、《系统科学与数学》、《应用概率统计》和《Econometrics》的副主编、编委或客座主编,是双法学会数据科学分会副理事长和国际统计学会当选会员。