A hybrid deep feature optimized ensemble framework for multiclass skin disease classification
Scientific Reports, Published online: 09 August 2026; doi:10.1038/s41598-026-65115-1 A hybrid deep feature optimized ensemble framework for multiclass skin disease classification
Skin conditions are among the most common dermatological disorders globally, and their accurate early diagnosis is crucial to minimizing diagnostic delays and associated mortality. This study introduces a new hybrid framework named ResBR-XGBR to classify skin lesions using HAM10000 dataset. The proposed method combines deep feature extraction with ResNet50 and EfficientNet-B0 models, followed by feature optimization using Boruta, Chi-square, and Recursive Feature Elimination (RFE) techniques to select the most discriminative features.
Hyperparameter tuning was performed on 80:20 training-testing split datasets, employing Grid Search Cross validation for optimal learning. The ResBR-XGBR ensemble framework outperformed individual classifiers, attaining a remarkable 95.2% test classification accuracy.
Written by urgent.news from Scientific Reports's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
