LAUSR

Coronary heart disease leads to a high mortality rate worldwide. Owing to delays in its detection, its treatment becomes challenging with little chances of recovery in many cases. An efficient, early-stage detection method is therefore urgently needed. Using the Framingham Heart Study Dataset, this study shows how data pre-processing via the multilayer perceptron following a deep learning approach will improve data quality when computing the likelihood of one having coronary heart disease. Apart from being highly efficient, our proposed approach results in highaccuracy of 96.50%. Finally, the paper discusses the rise in efficiency and accuracy achieved via use of deep learning techniques to enhance predictive outcomes v. traditional ones. The proposed study attempts to detect Coronary Heart Disease at an early stage.