Cost-Sensitive Neural Network for Prediction of Hypertension Using Class Imbalance Dataset

Abstract

Hypertension is a serious medical condition that significantly increases the risk of chronic diseases. Early detection of individuals at risk for hypertension allows to prevent and delay the incidence of related diseases and strokes. In recent years, numerous researches have been focused on the decision support system for predicting hypertension. However, the class imbalance has commonly occurred problem in real-world applications. In this paper, we present the end-to-end costsensitive neural network (COST-NN) framework incorporated with a weighted random forest-based feature selection technique to predict hypertension among Korean adults. First, it identifies the best representative features using a weighted random forest-based feature selection technique. Then, we apply the COST-NN for predicting target among hypertensive and non-hypertensive individuals. In order to identify the accurate predictive model, we compare the various baseline models. Experimental results showed that COST-NN outperforms the regular state-of-theart baseline models. In addition, the presented framework is expected to apply not only hypertension but also can support to prevent the patients from the risk of various diseases.

Explainable Artificial Intelligence Based Framework for Non-Communicable Diseases Prediction

“…Depends on the collected NCDs data in this study, we did not consider the frequently occurred class imbalance problem in prediction analysis during the experiment. Nonetheless, decision-making responses suffer from the class imbalance problem that also has received much attention from researchers [26,58]. To deal with this problem, sampling techniques are investigated to rebalance an imbalanced dataset to alleviate the effect of the skewed class distribution.…”

“…Various studies have focused on the accuracy enhancement of NCDs diagnostic models concerning feature selection techniques and refined machine-learning classifiers [18][19][20][21][22][23][24][25][26][27][28][29].…”

“…As determined machine learning techniques for NCDs, the wrapper, embedded or hybrid feature-selection techniques have been adopted more efficiently to select important feature subsets than common filtered techniques [26][27]. Moreover, it should be noted that most researchers in previous studies aimed to apply the feature selection techniques for not only higher accuracy but also an improvement in understanding the causes of NCDs.…”