Abstract: Data mining for healthcare is an interdisciplinary subject of research that has its roots in database statistics and may be used to assess the efficacy of medical treatments.. Diabetes is a chronic disease in which the pancreas fails to make adequate insulin or body fails to utilize the insulin that is produced appropriately. In the health-care business, data analysis plays a critical role in illness identification. The proposed research is being carried out to compare the performance of various classifiers in the Ada-Boost learning environment. We employed three distinct algorithms in this regard: BPN (back propagation neural network), SVM (support vector machine), C4.5 decision tree, and classifier. The ada-boost learning approach is used to train all of the algorithms. The diabetic disease type dataset from the UCI machine learning data repository is utilized in CSV format for training and testing developed classifiers. Hence, we implemented proposed hybrid classification system using the JAVA WEKA machine learning library. The performance of the system for all the classifiers is calculated and compared in terms of Accuracy, Error Rate, Time and Memory usages based on various experiments and datasets. Furthermore we also compared our system to the traditional base Decision Stump method.
Keywords: Data Mining, Machine Learning, Classification, Dataset, ensemble learning, boosting, Ada-Boost, diabetes disease, Prediction.
|
DOI:
10.17148/IJARCCE.2022.11149
