Multi-granularity decision rough set attribute reduction algorithm under quantum particle swarm optimization

The existing attribute reductions are carried out using equivalence relations under a complete information system, and there is less research on attribute reductions of incomplete information systems with new theoretical models such as multi-granularity decision rough sets. To address the above shortcomings, this paper first makes up a pessimistic-optimistic multi-granularity decision rough set model based on tolerance relations in incomplete information systems. The concepts of attribute importance and approximate distribution quality are introduced into the model to form an attribute reduction algorithm under incomplete information systems. Secondly, due to the NP-hard problem of attribute reduction, in order to further ensure the accuracy of the reduction result, this paper proposes a pessimistic-optimistic multi-granularity reduction algorithm under quantum particle swarm optimization. Experimental results on multiple-attribute data proved that the algorithm proposed in this paper can effectively attribute reduction in the decision table with missing data. At the same time, the algorithm of this paper has the role of iterative optimization search, ensuring the accuracy of the reduction results and increasing the applicability of multi-granularity decision rough sets.