Harris Hawks Feature Optimization for Identifying the Informative Pathogens of Pediatric Sepsis

One of the most fatal potentially life-threatening medical condition that increases the mortality in pediatric populations is pediatric sepsis. Unfortunately, the improper control of such disease can lead to tissue damage and organ dysfunction because of the overwhelming the human body's response to an infection. Therefore, early recognition and intervention can clearly improve outcome for infants and children with conditions that lead to sepsis before the admission to the intensive care unit (ICU). Accordingly, 17 informative differential expressed genes have been selected using a nature-inspired metaheuristic optimization algorithm known as Harris Hawks Optimization (HHO) for the early prediction of pediatric sepsis and evaluating their survival rate. Support vector machine (SVM) with Radial basis function (RBF) optimized using 20 runs of 5-fold cross validation proved the validity of the proposed model with an estimated average accuracy 96.64% to be used as a potential diagnostic biomarker for pediatric sepsis classification.