Abstract. Nonlinear component analysis is a popular nonlinear feature extraction method. It generally uses eigen-decomposition technique to extract the principal components. But th...
—Some data mining problems require predictive models to be not only accurate but also comprehensible. Comprehensibility enables human inspection and understanding of the model, m...
Abstract— This paper presents the use of a newly created network structure known as a Self-Delaying Dynamic Network (SDN). The SDNs were created to process data which varies with...
Visual information is processed in the posterior parietal cortex for the hypothesized purpose of extracting a variety of affordances for the generation of motor behavior. The term...
Automatic lipreading is automatic speech recognition that uses only visual information. The relevant data in a video signal is isolated and features are extracted from it. From a s...