Technology can be defined as an instrument which allows improved understanding medical data and better management of their health records. Rapidly changing medical technology and changing practice pattern of physicians have revolutionized health care monitoring. Today's medical research could be more advanced, more effective for the society by the application of computer algorithms in large medical data analysis. There is an ever increasing demand for technology based diagnostic predictions to anticipate and prevent complications of major diseases like diabetes, cancer, hypertension, and heart and liver disorders. Clustering is one of the data mining techniques for analyzing such medical datasets. It is a technique for finding similarity groups in data, called clusters. It groups data instances that are similar to each other in one cluster from the data instances that are very different from each other into a different cluster. Clustering is classified as an unsupervised learning task. The paper identifies different symptoms in different types of diabetic patients which are the clusters to be grouped. Similarity measure technique isolates the probable disease group for a particular patient. Fuzzy clustering extends this notion to associate each pattern with every cluster using a membership function. The paper finally predicts the most probable type of diabetes pertaining to a patient using the Minkowski metric of the unsupervised algorithm from the cluster of assumed symptom levels over a range. The same technique can be applied to predict any type of disease given the symptoms.
Cluster, diabetes, membership function, Minkowski metric, symptoms, data, gestational, matrices
