A computationally efficient approximation calculation method in dominance-based rough set approach

Recently Dominance-based Rough Set Approach (DRSA) has become a prominent tool for processing the datasets with preference order. However, DSRA requires calculating lower and upper approximations, which are computationally expensive. We have proposed a computationally efficient approximation calculation method that traverses each item in the dataset and updates the lower and upper approximations according to their preference order. In this way, there is no need to traverse the dataset multiple times, which not only enhances the performance but also produces the same results as produced by the conventional approach. Results on ten benchmark datasets from UCI have shown that the proposed approach significantly reduces the average computation time, that is, 19.5% in the case of lower approximation and 11% in the case of upper approximation. Memory consumption is also reduced by 99%. This clearly shows the efficiency and effectiveness of the proposed approach.