Author : Miss.Niphade Snehal Prabhakar

In this work, an effective scientific advice device that makes use of a quick Fourier transformation-coupled gadget studying ensemble model is proposed for short-time period disease danger prediction to provide persistent heart disorder patients with appropriate recommendations approximately the want to take a medical check or not on the approaching day based totally on reading their clinical information. The enter series of sliding home windows based at the patients time collection statistics are decomposed by way of the usage of the short Fourier transformation so one can extract the frequency facts. A bagging-based totally ensemble model is utilized to predict the patient’s condition someday in advance for generating the final recommendation. A combination of 3 classifiers synthetic neural community, least squares-help vector gadget, and naive bayes are used to construct an ensemble framework. A actual-lifestyles time series tele health facts gathered from chronic heart disease patients are applied for experimental evaluation. The advise device yields excellent recommendation accuracy and offers an powerful way to lessen the threat of wrong hints in addition to reduce the workload for coronary heart sickness sufferers in carrying out frame assessments every day. The proposed system is a promising tool for reading time collection medical statistics and providing correct and dependable pointers to sufferers laid low with persistent heart diseases.

 Keywords: Time Series Analysis, Intelligent Systems, Medical Data, Heart Disease, Recommender Systems