Incremental learning aims to alleviate the catastrophic forgetting problem of deep neural networks during sequential data stream. This is even more challenging when old unavailable, since system can only be trained under supervision current data. To address this problem, we proposed a prototype sampling mechanism based on K-means clustering method. On one hand, use pick out class-representative...

This paper presents two prototypes designed in order to enable the automatic adjustment of audio output on mobile devices. One is directed to communication scenarios and the other to media control scenarios. The user centered methodology employed on the design of these prototypes involved 26 users and is also presented here. Once the prototypes were implemented, a usability study was conducted....

Intelligent behavior is characterized by flexible and creative pursuit of endogenously defined goals. It has emerged in humans through the stages of evolution that are manifested in the brains and behaviors of other animals. Intentionality is a key concept by which to link brain dynamics to goal-directed behavior. The archetypal form of intentional behavior is an act of observation through time...

An examination of recent work on prototype-based architectures for object-oriented programming raises a number of interesting questions, the most obvious being: Are prototype-based object-oriented architectures superior to class-based architectures? Another way of addressing the same issues might be to ask: Are class-based architectures over-centralized, and excessively rigid? This work also ra...

The structural hallmark of retroviral transforming onc genes is a specific RNA sequence that is unrelated to the essential retroviral genes but closely related to certain cellular prototypes termed proto-onc genes. Two types of onc genes have been distinguished. Type I are onc genes which utilize elements of specific sequences only to encode a transforming protein. Type I1 onc genes are hybrids...

Hong Yan Department of Electrical Engineering, University of Sydney, NSW 2006, Australia Phone: +61 2 692-3515, fax: +61 2 692-3847, e-mail: [email protected] Abstract In this paper we present a method for designing optimized nearest neighbor classi ers. In this method, we produce initial prototypes for classi cation based on self-organization and optimize the prototypes using a multi-layer neura...

................................................................................................................................ 1

Software and other technical products offered to a mass market have a high demand on support and help desks. A tool for automated classification of incident reports, errors and other customer requests which offers previous (successful) hints or solution procedures could efficiently decrease support costs. We propose an approach to mining incidents and other customer requests for support based o...

The three systems described in this essay each target the most stringent security standards embodied in the A1 requirements as defined by the US Department of Defense " Trusted Computer Systems Evaluation Criteria " [DOD85], which is commonly called the Orange Book. While the initial designs of all these systems — and in some cases the implementations — predate the " Trusted Database Management...