Abstract: The problem of online multitask learning for solving multiple related classification tasks in parallel, aiming to classify every sequence of data received by each task accurately and efficiently. First, to meet the critical requirements of online applications which is highly efficient and scalable classification, this solution can make immediate predictions with low learning cost is needed. Second, classical classification methods, it is the batch or online, often encounter a dilemma when applied to a group of tasks i.e., a  process of single classify model trained on the entire collection of data from all tasks may fail to capture characteristics of individual task; on other process, a model trained independently on individual tasks may suffer from the insufficient training data. To overcome these kinds of challenges we have proposed a collaborative online multitask learning method, which will learn a global model over the entire data of all tasks.


Keywords:  Multitask learning, Global Model, collaborative Model, Scalable Classification.