This document summarizes a tutorial on multi-task learning. The tutorial aims to provide an understanding of multi-task learning concepts, approaches to modeling task relatedness, applications, and a multi-task learning software package. The tutorial roadmap covers multi-task learning background and motivations, formulations, case studies, and the MALSAR package. Methods discussed include mean-regularized multi-task learning, joint feature learning, trace-norm regularization, and modeling tasks as groups, graphs or trees. Applications include web page categorization, HIV therapy screening, and disease prediction.