LAUSR

As an important research topic of machine learning, multiclass classification has wide applications ranging from computer vision to bioinformatics. A variety of multiclass classification algorithms with promising performance have been proposed. Among them, the decomposition-based algorithms have shown their competitiveness, since they transform the original problem into several easily solved binary classification sub-problems. Unlike existing decomposition-based algorithms which tackle each sub-problem independently, this paper suggests an evolutionary multitasking method, named EMT-MC, for multiclass classification, where the concept of multitasking is introduced to achieve the multiclass classifier with better quality. To be specific, in EMT-MC, each binary classification sub-problem is firstly viewed as a task. Then, during the evolution, the tasks with low performance (termed “ill-solved” tasks) are aided by some well-selected “assisting” tasks by using the evolutionary multitasking learning. This not only ensures that the useful information in “assisting” tasks can be transferred into those “ill-solved” tasks, but also helps them to achieve classifiers with higher accuracy. Numerical experiments on different multiclass classification datasets demonstrate the superiority of the proposed method over the state-of-the-art algorithms.