CAFE Framework - Biosignal Data for Elderly Monitoring

CAFE Framework - Complete Raw Data Package for F1000Research This dataset contains the complete raw underlying data for the manuscript:"Compact Adaptive Feature Extraction (CAFE): A Bandwidth-Aware Multi-Domain Monitoring Framework for Elderly Care" Published in: F1000Research (2025)Authors: Abdulhameed ZN, Al Mashhadany Y, Salman TM FILES INCLUDED (13 files): Core Data Tables:- feature_statistics.csv - Population feature statistics (Table 3)- method_comparison_results.csv - Performance comparison (Table 4)- event_detection_results.csv - Event detection summary (Table 5)- bandwidth_adaptation_data.csv - Adaptive transmission (Table 7) Statistical Analysis:- bootstrap_resampling_results.csv - 1000 bootstrap iterations (ΔF1, 95% CI, p-value)- fall_risk_validation.csv - Binary classification metrics- roc_curve_data.csv - ROC curve coordinates (Figure 5, AUC=0.70) Reproducibility Files:- per_window_predictions_187.csv - All predictions for independent verification- fig3a_feature_distributions.csv - Figure 3a boxplot data- fig3b_feature_stability_cv.csv - Figure 3b stability data- fig3c_feature_importance.csv - Figure 3c importance rankings- analysis_parameters.txt - Complete methodological parameters- README.txt - Comprehensive documentation DATASET CHARACTERISTICS:- 163 subjects (128 EEG, 35 EMG/ECG/EOG)- 77 multimodal analysis windows (4-s, 50% overlap)- 187 labeled validation windows- Sampling rate: 256 Hz KEY RESULTS:- CAFE: 70.8% F1 (5 features, 0.040 kbps)- Statistical: 69.8% F1 (8 features, 0.064 kbps)- ΔF1 = 0.93 pp, 95% CI [-4.51, 6.43], p = 0.393- Bandwidth reduction: 37.5% LICENSE: CC BY 4.0 CITATION:Abdulhameed ZN, Al Mashhadany Y, Salman TM (2025). Compact Adaptive Feature Extraction (CAFE): A Bandwidth-Aware Multi-Domain Monitoring Framework for Elderly Care. F1000Research.