equation(7)XP= xjkfor1≤j≤s,1≤k≤n thevalueforxjkispresentinxkfor1≤k≤n be the set of the available attribute values, and
equation(8)XM= xjkfor1≤j≤s,1≤k≤n thevalueforxjkismissingfromxkfor1≤k≤n be the set of the missing attribute values.
2.2.1. Whole data strategy
The whole data strategy (WDS) XL019 a simple method for clustering incomplete data. When the proportion of incomplete data is small, WDS simply deletes all incomplete data and applies directly the standard FCM to the remaining complete data subset XW. In the WDS algorithm, the prototypes and the memberships of the data vectors in XW can be calculated by using the alternating optimization of ( 3) and (4), and the memberships of the data vectors in X−XWX−XW can be estimated by using a nearest-prototype classification scheme based on the following partial distance from each incomplete datum to each of the computed cluster prototypes
equation(9)Dik=s∑j=1sIjk∑j=1s(xjk−vji)2Ijk,where
equation(8)XM= xjkfor1≤j≤s,1≤k≤n thevalueforxjkismissingfromxkfor1≤k≤n be the set of the missing attribute values.
2.2.1. Whole data strategy
The whole data strategy (WDS) XL019 a simple method for clustering incomplete data. When the proportion of incomplete data is small, WDS simply deletes all incomplete data and applies directly the standard FCM to the remaining complete data subset XW. In the WDS algorithm, the prototypes and the memberships of the data vectors in XW can be calculated by using the alternating optimization of ( 3) and (4), and the memberships of the data vectors in X−XWX−XW can be estimated by using a nearest-prototype classification scheme based on the following partial distance from each incomplete datum to each of the computed cluster prototypes
equation(9)Dik=s∑j=1sIjk∑j=1s(xjk−vji)2Ijk,where