Title :
Improved Constructive Semi-Supervised Classification Algorithm for handling the overlapping of hyperspheres
Author :
Chandel, Arvind Singh ; Tiwari, Anish ; Patel, O.P.
Author_Institution :
Dept. of Comput. Eng., OPJIT, Raigarh, India
Abstract :
An Improved Constructive Semi-Supervised Classification Algorithm (ICS-SCA) is proposed to handle the problem of overlapping among hyperspheres of different classes. The overlapping problem occurs in Constructive Semi-Supervised Classification Algorithm proposed in [1]. Unlike CS-SCA[1], approach proposed here handles the samples lying in overlapping region of hyperspheres belonging to the different classes while it allows the overlapping among hyperspheres belonging to the same classes. Proposed approach is tested with the datasets that were taken in CS-SCA[1] and improved results have shown in table.
Keywords :
neural nets; pattern classification; ICS-SCA; binary neural network; hypersphere overlapping; improved constructive semisupervised classification algorithm; Accuracy; Breast; Cancer; Radio frequency; Unsupervised learning; Binary Neural Network; Geometrical Expansion; Hyperspheres (HS); Overlapping; Semi-Supervised Classification;
Conference_Titel :
Issues and Challenges in Intelligent Computing Techniques (ICICT), 2014 International Conference on
Conference_Location :
Ghaziabad
DOI :
10.1109/ICICICT.2014.6781265