Enhancing the Accuracy for Predicting Nutritious Food Recommendations for Patients using Ensemble Random Forest over Artificial Neural Network
J Mukund, A. Akilandeswari
Abstract
This study uses Innovative Ensemble Random Forest (ERF) and Artificial Neural Network to propose healthy diet to patients based on their health. Data collection and model training are required. This study uses Innovative Ensemble Random Forest (ERF) and Artificial Neural Network (ANN) data mining approaches on two groups. Each group has 10 samples for 20 total. Sample size is calculated using 0.95 G power. The parameters are CI 95, alpha 0.05, and beta 0.002. ERF and ANN should be used to train and assess healthy meal recommendation systems. The Innovative Ensemble Random Forest (ERF) approach outperforms the Artificial Neural Network algorithm with a 95.72 % accuracy and a statistical significance of 0.008 (p < 0.05).The proposed research's accuracy comparison with earlier convolutional techniques and nutrient-dense meal recommendation demonstrate its excellence. Innovative Ensemble Random Forest is more accurate than Artificial Neural Network with ERF 95.72 % and ANN 94.38 %.