Subject variability in sensor-based activity recognition

• Published in: Journal of ambient intelligence and humanized computing Vol. 14; no. 4; pp. 3261 - 3274
• Main Authors: Jimale, Ali Olow, Mohd Noor, Mohd Halim
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.04.2023; Springer Nature B.V
• Subjects: Activity recognition; Age; Age differences; Age groups; Artificial Intelligence; Artificial neural networks; Caregivers; Classification; Classifiers; Cognitive ability; Computational Intelligence; Data collection; Datasets; Decision trees; Deep learning; Engineering; Inertial sensing devices; Investigations; Life expectancy; Machine learning; Older people; Original Research; Performance degradation; Robotics and Automation; Sensors; Support vector machines; Test sets; User Interfaces and Human Computer Interaction; Deep learning; Activity recognition; Subject variability; Machine learning

Building classification models in activity recognition is based on the concept of exchangeability. While splitting the dataset into training and test sets, we assume that the training set is exchangeable with the test set and expect good classification performance. However, this assumption is invali...