期刊名称：International Journal of Innovative Research in Science, Engineering and Technology
印刷版ISSN：2347-6710
电子版ISSN：2319-8753
出版年度：2017
卷号：6
期号：6
页码：12329
DOI：10.15680/IJIRSET.2017.0606306
出版社：S&S Publications
摘要：Classification and clustering algorithms were used separately in prediction of Thyroid Disease. Successand accuracy achieved by classification and clustering algorithm for thyroid prediction was less. We can take advantageof both classification and clustering algorithm for thyroid disease prediction by combining them in system. This paperfocuses on potential benefit of using artificial neural networks (ANNs) as a classification algorithmand improvised Kmeansas clustering algorithm for the diagnosis of thyroid.We have examined the ANN architecture and assessed theirrobustness in the face of diagnostic noise. The experimental dataset of thyroid has previously studied by multivariatestatistical methods and a variety of pattern-recognition techniques. For the experimentation, the thyroid dataset isdivided into two subsets, one is training dataset for the networks and another for testing network performance. The testdataset has various proportions of cases with diagnostic noise to mimic real-life diagnostic situations.
