DCD_PREDICT: Using Big Data on Prediction for Chest Diseases by Applying Machine Learning Algorithms

Abstract

Technology with system learning algorithms is frequently utilized in the medical domains to estimate disorders. Through the provision of some reference guidelines, it assists in the real-world diagnosis of diseases. The DCD-PREDICT system employs system learning to make prophetic diagnosis of diseases of the chest, including lung cancer, asthma, COPD, pneumonia, and tuberculosis. A questionnaire will be provided to each participant (self-administered and physician-administered). Understanding, specificity, and positive and negative analytical values will be computed for each question, and the combined patient scores will be contrasted with those of controls. It will be determined how closely the physician- and self-administered questionnaires agree. This enables medical professionals to do better differentiated analysis earlier, lowering errors and delivering timely treatment. One of the main causes of death can be the heart disease. Because real-world practitioners lack the necessary knowledge, expertise, or experience regarding the signs of heart failure, it is challenging to diagnose the disease. Therefore, computer-based predictions of cardiac illness may be crucial as an early diagnosis to take the appropriate actions as well as a perspective on recovery. However, by choosing the right data mining classification algorithm, the early stages of the disease and its recurrence can be accurately predicted. The aim of this study was to compare three of the most common classification methods, Support Vector Machines (SVM), K-Nearest Neighbors (KNN) and Artificial Neural Networks (ANN), for heart disease prediction using the ensemble of standard Cleveland cardiology data.