一种融合模糊覆盖的模糊概念认知学习
打开文本图片集
中图分类号：TP182 文献标志码：A 文章编号：1673-2340（2025）03-0023-11
A fuzzy concept-cognitive learning method integrating fuzzy coverage
WUYuqing1 23，LINYidong1，2，3，LIANG Taoju1，2，3 （l.School of Mathematics and Statistics，Minnan Normal University，Zhangzhou 363ooo，China; 2.Institute of Meteorological BigData-Digital Fujian，Zhangzhou 363ooo， China; 3.Fujian Key Laboratory of Granular Computing and Applications， Zhangzhou 363ooo， China）
Abstract：Concept-cognitive learning isan emerging interdisciplinaryresearch area that aims tocontinuously learn new knowledgebyimitating the human cognitive process.However，existingconcept-cognitive learning models usuallyignore the local variabilityof objects inconcepts，theredundancyof concept space，andthe interpretabilityof concepts，which leads to modelcognitive biasand underutilization ofvalid information.Therefore，a fuzzyconcept-cognitive learning model integrating membership degree and coverage is proposed in this paper.Firstly，to enhance the representation capabilityof theconcept extension，amembership function withanoffsetthreshold is introduced to explore the corelation between objects and the concept.Amembership matrix is then constructed to furthertransform the concept space into a fuzzy coverage.Secondly，high-correlation objects are filtered through the fuzzy β cut set，and the importance of different concepts is explored through coverage rates.This enables the constructionof acore concept space，which effectively reduces theredundancy of the concept spaceand enhancescognitive learning effciency. Finally，the proposed modeliscomparedwith four machine learningalgorithms andtwoconcept-cognition algorithms usingthe ten-foldcrossvalidationmethod.Theexperimentalresults demonstrate thatthemodelachieves higheraverage accuracy than the other comparative algorithms across l4 datasets，with thesmalest performance fluctuation rangeacross different datasets.Moreover，itmaintainsa leadingpositionin termsof precision，recalland F1 score， fullyvalidating the feasibilityand effectiveness of the proposed model.
Key words： concept-cognitive learning; membership matrix; fuzzy coverage; core concept space
概念认知学习（concept-cognitivelearning，CCL）是通过概念进行事物认知与学习的科学川。（剩余14983字）
