Unlabeled

This is the source code and script from the case study presented during the Data Analysis Webinar on 4/24/03 (archived at http://www.mathworks.com/cmspro/online/5102/req.jsp?13727). The case study demonstrates how several MATLAB toolboxes can be used together to handle a real-life engineering... Platforms: Matlab

This Python package contains some variants of the matching pursuit sparse coding algorithm. Matching pursuit uses a set of "basis functions" or "codebook filters" to greedily encode a raw signal in terms of a weighted sum of filters. Using gradient ascent we can also infer a likely set of filters... Platforms: *nix

parameter 1 is feature value (discrete), parameter 2 is the corresponding labelsthe reference paper is: S. Wu, P. A. Flach, Feature selection with labeled and unlabeled data, available athttp://ai.ijs.si/branax/iddm-2002_proceedings/Wu.pdfrelevant paper is: Technical Note: Bias in... Platforms: Matlab