In typical classification problems, high level concept features provided by a domain expert are usually available during classifier train ing but not during its deployment. We ad dress this problem from a multitask learn ing (MTL) perspective by treating these fea tures as auxiliary learning tasks. Previous efforts in MTL have mostly assumed that all tasks have the same input space. How ever, auxiliary tasks can have different input spaces, since their learning targets are dif ferent. Thus, to handle cases with heteroge neous input, in this paper we present a newly developed model using heterogeneous auxil iary tasks to help main task learning. First, we formulate a convex optimization problem for the proposed model, and then, we ana lyze its hypothesis class and derive true risk bounds. Finally, we compare the proposed model with other relevant methods when ap plied to the problem of skin cancer screening and public datasets. Our results show that the performance of the proposed method is highly competitive compared to other rele
【 预 览 】
附件列表
Files
Size
Format
View
Assisting Main Task Learning by Heterogeneous Auxiliary Tasks with Applications to Skin Cancer Screening