Prediction of Child Malnutrition Status using Neutrosophic Logic, Feature Ranking and Optimization Techniques in Deep Convolutional Neural Network

Problem: Today child growth, in future can lead the country in economic development. The child malnutrition problem can lead the society in under development and minimize the child growth in future. The malnutrition issues can create more problems in their daily life. Nutrition is essential factor for child growth. In India, the child malnutrition problem can be increased year by year. Various factors like poor nutrition intake, economic problems and social factors can cause child malnutrition problem. In the previous methodology the uncertainty issues cannot be resolved. To address this problem, this study was developed using deep learning method with convolutional neural network and implement neutrosophic set domain, feature ranking algorithm and optimization techniques to improve the accuracy in child malnutrition prediction. This research work used Bipolar Neutrosophic Set (BNS), feature ranking and optimization algorithm to improve the accuracy. The deep learning algorithm of CNN- Convolutional Neural Network architecture has proposed to predict the child malnutrition. The outcome of this model can be measured in different metrics including Accuracy, Precision, Recall, and F1-score. For NFHS dataset, the dove swarm optimization produced accuracy 97.48%, Precision 98.02%, Recall 97.15% and F1-score 98%. For National Nutritional Survey dataset, optimization algorithm produced accuracy 98.58%, Precision 97.46%, Recall 98.11% and F1-score 97.45 %. For CM-UNICEF Dataset, the optimization method produced accuracy 98.16%, Precision 97.15%, Recall 98.50% and F1-score 98.26 %. Optimization method is better for prediction compared to bipolar neutrosophic set domain and feature ranking method. Comparison about the methods of bipolar neutrosophic set domain algorithm, feature ranking algorithm and dove swarm optimization technique, the optimization algorithm produced more accuracy which is recommended to predict the child malnutrition.