Segmenting information records with missing values using multiple partition trees

摘要: A method and system for predicting the class membership of a record where information one or more variables in is missing. Multiple classification trees are generated. first tree computed using substantially complete set all variables. Other different subsets Variables selected inclusion subset based on how strongly they influence prediction membership. The (based information) applied to with missing information. If needed by this order classify record, another that not variable selected. predicted accurately without increasing complexity prediction.