A generalized arbitrary base euclidean space technique for C-Mean clustering simulation
摘要
In this article, a resistant metric space with a Gaussian distance-dependent function is considered due to its novelty. New metrics provide rigid and concise initial criteria, which depends upon metric properties and results based upon experiments. Improved Alternative Hard C-Means (IAHCM) and Improved Alternative Fuzzy C-Means IAFCM were used to get strong results by reducing their sensitivity but now a substitute of arbitrary constant is attempted to enhance the performance of IAHCM and IAFCM. The major focus is clustering and the development of novel clustering algorithms and the terms that describe it are alternative generalized hard c-mean (AGHCM) and alternative generalized fuzzy c-mean (AGFCM). Two and high-dimensional data in our experiments are employed, including diamond data collection, Iris real-life data, and glass datum making the proposed system accessible, resilient and effective.