Missing value imputation using adaptive ordering and clustering analysis

摘要

As received, a data value of an expected input set of received data values is missing from user input. A subset of known data with data values similar to a subset of the received data values is determined. A data sample average for the missing data value is determined from data values within the subset of the known data. An initial estimate of the missing data value is initialized using the data sample average. Boundary data clusters near the initial estimate of the missing data value are identified within the subset of the known data. A data harvesting region encapsulated according to the boundary clusters is defined. Data support clusters within at least one subset of the known data inside the data harvesting region are selected. The initial estimate of the missing data value is updated based upon data of the boundary clusters and the data support clusters.