Estimating relevant labels Jointly based on multivariate regression

摘要: Recent advances have shown that some intrinsic traits of personality are reflected in the resting-state functional brain network (FBN), which opens a new way evaluating more objectively (relative to traditional questionnaire-based methods). However, current studies evaluate different factors independently without considering their potential relationship. To address this problem, we propose estimate several factor scores jointly for evaluation by introducing factor-similarity matrix can be automatically updated according FBN data. Additionally, an L 1 -norm regularizer is used estimating method obtain sparse solution toward better interpretability. As result, realize multi-variate linear regression, feature selection and factor-relationship learning unified framework. Finally, design alternating optimization algorithm solve proposed method. Experimental results on Human Connectome Project (HCP) dataset show method, with learned relationship, achieve higher estimated accuracy than related