A relevance filter is proposed which removes features based on the mutual information between class labels and features. It is proven that both feature independence and class condi...
Frequent itemset mining assists the data mining practitioner in searching for strongly associated items (and transactions) in large transaction databases. Since the number of frequ...
We outline the application of a framework for managing information quality (IQ) in proteomics. The approach allows scientists to define the quality characteristics that are of im...
Alun D. Preece, Binling Jin, Paolo Missier, Suzann...
The widespread adoption of the JPEG2000 standard calls for the development of computationally efficient algorithms to analyze the content of imagery compressed using this standard...
Ali Tabesh, Ali Bilgin, Karthik Krishnan, Michael ...
Abstract. Bayesian spam filters, in general, compute probability estimations for tokens either without considering the email areas of occurrences except the body or treating the s...