Utility Maximization

We consider the problem of training logis- tic regression models for binary classifi- cation in information extraction and infor- mation retrieval tasks. Fitting probabilis- tic models for use with such …...