In order to maintain their performance in a dynamic environment, agents may be required to modify their learning behavior during run-time. If an agent utilizes a rule-based system...
Stephen Quirolgico, K. Canfield, Timothy W. Finin,...
Abstract. In this paper we investigate a new method of learning partbased models for visual object recognition, from training data that only provides information about class member...
Typical content-based image retrieval (CBIR) solutions with regular Euclidean metric usually cannot achieve satisfactory performance due to the semantic gap challenge. Hence, rele...
The existing Multi-View Learning (MVL) learns how to process patterns with multiple information sources. In generalization this MVL is proven to have a significant advantage over...
In this research, we devised a new simple technique for statically holding analog weights, which does not require periodic refreshing. It further contains a mechanism to locally u...
Bassem A. Alhalabi, Qutaibah M. Malluhi, Rafic A. ...