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...
Scientific workflow systems increasingly store provenance information about the module executions used to produce a data item, as well as the parameter settings and intermediate...
Susan B. Davidson, Sanjeev Khanna, Tova Milo, Debm...
In this paper, we consider the problem of combining link and content analysis for community detection from networked data, such as paper citation networks and Word Wide Web. Most ...
Background: Profile hidden Markov model (HMM) techniques are among the most powerful methods for protein homology detection. Yet, the critical features for successful modelling ar...
Previous pixel-level change detection methods either contain a background updating step that is costly for moving cameras (background subtraction) or can not locate object positio...