HYPERTENSION PREDICTION USING DEEP LEARNING WITH TRANSFER LEARNING TECHNIQUES

Abstract

Hypertension or high blood pressure is a chronic condition of consistent rise in blood pressure above the identified normal. It significantly increases the risk of cardiovascular diseases when identified at an advanced stage, but when diagnosed and treated early, it reduces the occurrence of life-threatening complications. This research proposes a prediction model using Deep Learning (DL) with Transfer Learning (TL) techniques for early prediction of hypertension. A pre-trained Feed-Forward Deep Neural Network model, initially developed for diabetes prediction using the PIMA diabetes dataset, is fine-tuned for hypertension prediction using the PPG-BP dataset. This approach utilizes the model's ability to transfer learned knowledge, improving accuracy while reducing computational time. The performance of the model is evaluated using accuracy, precision, and recall. It achieved an accuracy of 81.34%.

Ähnliche Arbeiten