Answers are grounded in this paper and cite their sources.
Accurate estimation of body fat percentage (BFP) is essential for obesity assessment and cardiovascular risk management. However, conventional anthropometric methods lack sufficient precision, while specialized body composition analyzers entail high costs that limit their widespread clinical adoption. This study proposes a multi‑objective metaheuristic‑driven feature selection framework that integrates an artificial neural network (ANN) with the Non‑Dominated Sorting Genetic Algorithm II (NSGA‑II) to optimize both predictive R2 and feature parsimony for BFP estimation. The proposed methodology is evaluated on a benchmark dataset comprising 252 individuals with 13 anthropometric predictors. A comprehensive baseline assessment is conducted using 16 machine learning algorithms, including six linear and ten non‑linear models. The hybrid ANN‑NSGA‑II framework identifies an optimal subset of six features, achieving a root mean square error (RMSE) of 3.4962 and an R‑squared (R²) value of 0.8311. These quantitative outcomes demonstrate that embedding a multi‑objective metaheuristic selection mechanism within an ANN architecture enables substantial dimensionality reduction while maintaining robust predictive performance. The proposed approach provides a cost‑effective and computationally efficient alternative for BFP prediction using only fundamental anthropometric measurements, thereby eliminating dependence on specialized body analysis equipment.
No takes yet. Share an insight, caveat, or question.
Yousefi et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: