Research Article
Parallel Attribute Reduction Algorithm for Complex Heterogeneous Data Using MapReduce
Input: a data split Si | Output: <KEYUM, VALUEUM> | // key^ represents the sample does not belongs to the positive region; value^ represents the sequence of the sample’s attribute value | begin | <KEYUM, VALUEUM>= | for each do | if key^i=1, value^i=xi | else key^i=0, value^i=xi | end if | <KEYUM, VALUEUM>=<KEYUM, VALUEUM> <key^i, value^i> | end for | end |
|