A Semi-Supervised Deep Fuzzy C-Mean Clustering for Two Classes Classification
Deep Fuzzy C-Means algorithm is applied to determine veiled structure in the data set.It is commonly used when data boundaries are not clearly defined and extra parameters are needed to reduce the statistical closeness.In this paper,we propose a semi-supervised deep fuzzy C-Means algorithm that accommodates this intangibility.It is applicable to machine learning methodology that relies on algorithmic flow for dynamic data.With statistical data provided in the form of a collection of numerical data set of two classes,namely labeled and unlabeled,the semi-supervised deep fuzzy c-means clustering provides a comparison and solution for a given data set.The clustering approach looks at membership functions for fuzziness.The proposed framework for semi-supervised data set finds supervised data and segregates it from unsupervised data.Here,the term “deep defines the proximity in space,which is used to improve precision along the centers.The membership function for each cluster is used to gauge the closeness between the unlabeled and labeled data set.The dependency of our algorithms performance on control parameters helps us determine the variability of the clustering technique.Our simulation result shows that semi-supervised deep fuzzy-c algorithm performs better than previously studied semisupervised clustering algorithms.
Semi-supervised learning Supervised Learning Fuzzy C-Mean Deep Fuzzy C-Mean
Ali Arshad Saman Riaz Licheng Jiao Aparna Murthy
School of Computer Science and Technology and School of International Education,Xidian University,Xi Key Laboratoy of Intelligent Perception and Image Understanding of Ministry of Education of China
国际会议
重庆
英文
365-370
2017-10-03（万方平台首次上网日期，不代表论文的发表时间）