论文部分内容阅读
Computational models of learning are typically divided into supervised learning, unsupervised learning,Computational models of learning are typically divided into supervised learning, unsupervised learning,for expanding incomplete knowledge, via self-directed learning that incorporates knowledge not previously experienced. This article defines a new self-supervised learning framework to address these pregnant learning contexts, and implements this framework using adaptive resonance theory. The learning framework learns about novel features from unlabeled patterns without destroying knowledge previously acquired from labeled patterns.