Performance Measure of Classifier for Prediction of Healthcare Clinical Information

Abstract

In the healthcare sector, there are various types of patient data, and that data need to be preserved for the future diagnosis of that particular patient and such a large size data can be stored using a concept of big data. In healthcare services, a huge amount of healthcare information is regularly generated at a very high speed and volume. Traditional databases are unable to handle such a huge amount of data. Every day increasing the volume of digital health care information has providing new opportunities leads to the quality of health care services and also avoid the repeated medical tests cost. If all the healthcare information is available in the form of digital, then we can use various tools and technologies to process healthcare information and generate decisions regarding the prediction of disease. Our proposed system is the automated clinical decision support system in association with a classifier. An objective of the implemented system has to predict the disease, using various classification techniques. The healthcare raw information data are stored and features are extracted that are used in further processing; based on those features, analysis is done and generates decisions on patient health information which are supplied. The proposed system is followed by a pipelined architecture and it contains the following phases: storage, feature extraction, classification, analysis, searching, and decisions. Research work emphasis on multiple classification techniques to increase the accuracy of prediction of patient health information.