Trans-Glasso: A Transfer Learning Approach to Precision Matrix Estimation
Trans-Glasso:一种用于精准矩阵估计的迁移学习方法
机构 * Booth School of Business, University of Chicago(芝加哥大学布斯商学院) ; Department of Statistics, University of Chicago(芝加哥大学统计系) ; Department of Data Sciences and Operations, Marshall School of Business, University of Southern California(南加州大学马歇尔商学院数据科学与运营系)
AI总结 Trans-Glasso通过迁移学习方法提升小样本下的精准矩阵估计精度,结合多任务学习和微分网络估计,在基因网络和蛋白质网络中验证了其有效性,并推导了微分网络估计的最优率。
Comments 58 pages, 13 figures. Accepted by the Journal of the American Statistical Association (JASA)