Convergence of blind delayed source separation algorithms, which use constant learning rates, is known to be slow. We propose a fuzzy logic based approach to adaptively select the...
Discretization algorithms have played an important role in data mining and knowledge discovery. They not only produce a concise summarization of continuous attributes to help the ...
: This paper shows how the Asynchronous Backtracking (Yokoo et al., 1998) algorithm, a well known distributed constraint satisfaction algorithm, produces unnecessary messages and i...
The relative worst order ratio is a measure for the quality of online algorithms. Unlike the competitive ratio, it compares algorithms directly without involving an optimal offline...
: Document binarization is an active research area for many years. The choice of the most appropriate binarization algorithm for each case proved to be a very difficult procedure i...