Wavelet-Based Decision Tree Model with Entropy Index for Classifying Fall History Groups

This study investigates the classification of fall history using wavelet-based entropy features derived from center of pressure (COP) signals. Fall history data were collected from ۱۳۰ participants aged ۶۰-۸۱ years through detailed interviews and force plate measurements using standardized quiet standing protocols. Wavelet entropy analysis was applied to decompose COP signals into multiple frequency bands, followed by entropy calculation at each level and classification using an optimized decision tree algorithm. We achieved ۹۲% classification accuracy in distinguishing between fallers and non-fallers. This study highlights the potential of wavelet-based entropy measures for fall history classification and their clinical applications in fall risk assessment.

کلیدواژه ها:

Fall Risk ، Wavelet Entropy ، Decision Tree ، Machine Learning ، Center of Pressure