Optimizing Fuzzy System of Fuzzy Time Series for Hyper Spectral Image Classification

Abstract

This research paper examines the capability of fuzzy time collection for hyperspectral photograph classification. Fuzzy time series (FTS) is a time series in which fuzzy standards are used to model the styles within the facts. FTS can be used to explain complex temporal styles in the records, and as a consequence making it possible to categorize photographs more extraordinarily accurately., this look proposes an optimization method primarily based on genetic seek techniques. The optimization algorithm is designed to discover the high-quality FTS parameters that yield first-rate type accuracy. The efficacy of the proposed technique is evaluated on hyperspectral facts set with extraordinary experimental scenarios. The results of the test display that the proposed method can enhance the accuracy of photo classification and the use of FTS considerably. Hence, the proposed method gives a promising technique that can be used to classify hyperspectral snapshots efficiently. The paper affords an optimized fuzzy machine of fuzzy time collection for the hyperspectral photograph category. The proposed device consists of 3 levels: pre-processing, version creation, and optimization. Throughout the pre-processing level, statistical and spectral analyses are executed to acquire the applicable attributes for developing the fuzzy time collection. The model construction degree then uses the bushy time series to extract between-class separability for the photo type. It is followed utilizing the optimization stage, related to the software of differential evolution, to minimize the complexity of the proposed machine while still enhancing the type accuracy. The proposed machine has been correctly carried out to a real-international hyperspectral dataset and demonstrates widespread upgrades in class accuracy over existing methods.