Big Data Science in Building Medical Data Classifier Using Naïve Bayes Model

Abstract

currently, maintenance of clinical databases has become a crucial task in the medical field. The patient data consisting of various features and diagnostics related to disease should be entered with the utmost care to provide quality services. As the data stored in medical databases may contain missing values and redundant data, mining of the medical data becomes cumbersome. As it can affect the results of mining, it is essential to have good data preparation and data reduction before applying data mining algorithms. Prediction of disease becomes quick and easier if data is precise and consistent and free from noise. One of the key specialty of Naive Bayes classifiers is that they are highly scalable, requiring a number of parameters linear in the number of variables (features/predictors) in a learning problem. Evaluation of closed-form expression can be achieved by Maximum-likelihood training. Which requires linear time, rather than by expensive iterative approximation as used for many other types of classifiers. This research uses data science approach to diognize the medical data. In this article, a study has been conducted by using naïve Bayes classifier to classify the medical data. The suitability of the classifier and the accuracy of the classifier are measured using different performance criteria. This study is useful for researchers and developers in understanding and using a classification technique in medical diagnosis.