| C. Li. Extending ITERATE conceptual clustering scheme in dealing with numeric data. Master's thesis, Vanderbilt University, 1995. |
....cancer data sets were selected because they contain a high number of features (at least for the standards of the UCI Repository) Table 1 summarizes the main characteristics of each data set. Data sets containing numerical features have been discretized using the unsupervised method proposed by [16]. It is worth noticing at this point that the notion of irrelevance may be somewhat di erent for supervised and unsupervised learners. Although completely irrelevant features for supervised tasks such as those added in the LED and waveform data sets should be also irrelevant for clustering, it is ....
C. Li. Extending ITERATE conceptual clustering scheme in dealing with numeric data. Master's thesis, Vanderbilt University, 1995.
....Cancer data sets were selected because they contain a high number of features (at least for the standards of the UCI Repository) Table 1 summarizes the main characteristics of each data set. Data sets containing numerical features have been discretized using the unsupervised method proposed by Li (1995). It is worth noticing that, although completely irrelevant features for supervised tasks such as those added in the LED and Waveform data sets are likely to be also irrelevant for clustering, it could happen that redundant features in supervised environments were not considered as such for ....
Li, C. (1995). Extending ITERATE conceptual clustering scheme in dealing with numeric data. Master 's thesis, Vanderbilt University.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC