Ant Colony Optimization Toward Feature Selection
Monirul Kabir, Md Shahjahan, Kazuyuki Murase · InTech eBooks · 2013
Over the past decades, there is an explosion of data composed by huge information, because of rapid growing up of computer and database technologies. Ordinarily, this information is hidden in the cast collection of raw data. Because of that, we are now drowning in informa‐ tion, but starving for knowledge [1]. As a solution, data mining successfully extracts knowl‐ edge from the series of data-mountains by means of data preprocessing [1]. In case of data preprocessing, feature selection (FS) is ordinarily used as a useful technique in order to reduce the dimension of the dataset. It significantly reduces the spurious information, that is to say, irrelevant, redundant, and noisy features, from the original feature set and eventually retain‐ ing a subset of most salient features. As a result, a number of good outcomes can be expect‐ ed from the applications, such as, speeding up data mining algorithms, improving mining performances (including predictive accuracy) and comprehensibility of result [2].