Skip to main content
. Author manuscript; available in PMC: 2011 Jun 1.
Published in final edited form as: Pattern Recognit. 2010 Jun 1;43(6):2340–2350. doi: 10.1016/j.patcog.2009.12.003

Fig. 4.

Fig. 4

Performance of missing value management methods versus cluster validity index: (a)-(c) the results of the UCI Breast Cancer and Pima Diabet datasets, and the human brain image database system (HBIDS), respectively (refer to Table 1). The selection-fusion (SF), expectation maximization (EM), multiple imputations (MI), and CART methods are compared. 1/CVI increases as the percentage of missing values increases. Note the overall superiority of the SF method especially when 1/CVI is large.