Regret-Theoretic Multiattribute Decision-Making Model Using Three-Way Framework in Multiscale Information Systems

KNN Data Filling Algorithm for Incomplete Interval-Valued Fuzzy Soft Sets

As a generalization of the fuzzy soft set, interval-valued fuzzy soft set is viewed as a more resilient and powerful tool for dealing with uncertain information. However, the lower or upper membership degree, or both of them, may be missed during the data collection and transmission procedure, which could present challenges for data processing. The existing data filling algorithm for the incomplete interval-valued fuzzy soft sets has low accuracy and the high error rate which leads to wrong filling results and involves subjectivity due to setting the threshold. Therefore, to solve these problems, we propose a KNN data filling algorithm for the incomplete interval-valued fuzzy soft sets. An attribute-based combining rule is first designed to determine whether the data involving incomplete membership degree should be ignored or filled which avoids subjectivity. The incomplete data will be filled according to their K complete nearest neighbors. To verify the validity and feasibility of the method, we conduct the randomized experiments on the real dataset as Shanghai Five-Four Hotel Data set and simulated datasets. The experimental results illustrate that our proposed method outperform the existing method on the average accuracy rate and error rate.