Thyroid classification using Deep Learning Techniques

Thyroid is spreading all over India. About the disorder, this Study aims to provide the source of research to a clear prediction in thyroid classification. From the various machine learning techniques, four algorithms used. The algorithms are Decision tree, Naïve Bayes, Random Forest and Support Vector Machine (SVM) to evaluate and predict the performance accuracy. The study has highlighted the SVM algorithm to overcome with best accuracy using Hybrid. The data are collected from UCI Machine Learning Repository. Intensive Subset Cluster Feature Selection (ISCF) to choose the features depending on marginal accuracy. This reduces the big data dimensionality problems which make features to trained on Multi-Perceptron Neural Network (MPNN) for best training features. The proposed ISCF-MPNN resultant factors prove that best classification accuracy is achieved with regard to thyroid disease influence rate which is up to 97 %. on Decision tree and SVM feature with random forest classification (HDT-SVMRF. Finally, Optimized Thyroid classification and prediction based on Deep featured spectral multi perception neural network is implemented to improve the thyroid classification.