Classification of high-temperature superconducting YBCO thin films by fuzzy clustering

The present paper deals with the classification of high-temperature superconducting thin films (HTS) by the use of fuzzy clustering. A dataset from 84 YBCO thin films obtained by chemical vapour deposition technology was treated by the fuzzy approach. The films described were summarized by their input (oxygen pressure, chemical precursors, substrate parameters) and output (critical temperature, critical current) physical characteristics. The aim of the study was to confirm a previous classification of the films by crisp cluster analysis, and to improve the film separation into similar clusters by the more advanced and reliable fuzzy clustering method. It has been shown that fuzzy clustering allows a better classification, ascribing a membership function to each HTS thin film.