A Survey on Prediction of Diabetes using Data Mining

Abstract

With recent advances in computer technology, large amounts of data will be collect and store, but all this data becomes more useful when it is analyze and some dependencies and correlations has found. This should be an accomplished with machine learning algorithms. The automatic diagnosis of diabetes is an important real-world medical problem. Detection of diabetes in its early stages is the key for treatment. A study of existing systems reveals that it is important to discover the hidden knowledge from a particular dataset to improve the quality of health care for diabetic patients. Techniques used in existing systems are namely: EM algorithm, H-means+ clustering and Genetic Algorithm, for the classification of the diabetic patients. In addition, Ant Colony Optimization (ACO) has used in data mining field to extract rule based on classification systems. Survey work shows that there are data mining algorithms which can be used in the analysis to develop a more efficient prediction technique.