We propose a unified manifold learning framework for semi-supervised and unsupervised dimension reduction by employing a simple but effective linear regression function to map the ...
Feiping Nie, Dong Xu, Ivor Wai-Hung Tsang, Changsh...
This paper proposes three novel training methods, two of them based on the back-propagation approach and a third one based on information theory for Multilayer Perceptron (MLP) bin...
A secure logging mechanism named haplog is proposed. In essence, haplog uses only one-way hash functions to achieve the security and functional requirements of logging. This makes...
Typical approaches to multi-label classification problem require learning an independent classifier for every label from all the examples and features. This can become a computati...
This contribution describes a neural network that self-organizes to recover the underlying original sources from typical sensor signals. No particular information is required abou...