Abstract

This study aims to develop a robust metalearning system for rapid classification on a large number of tasks. The model-agnostic metalearning with the CACTUs method is improved as EW-CACTUs-MAML after integrated with the entropy weight method. Few-shot mechanisms are introduced in the deep network for efficient learning of a large number of tasks. The process of implementation is theoretically interpreted as “gene intelligence.” Validation of EW-CACTUs-MAML on a typical dataset indicates an accuracy of 97.42%, performing better than CACTUs-MAML. At the end of this paper, the availability of our thoughts to improve another metalearning system is also preliminarily discussed based on a cross-validation on another typical dataset.