Application of machine learning methods in flexible and wearable sensing systems

Flexible wearable devices provide essential technological support for the real-time and continuous monitoring of vital signs and motion signals. With the rapid advancement of functional materials and sensing technologies, various flexible wearable sensors have generated large volumes of complex multimodal data. Applying advanced machine learning algorithms from the field of artificial intelligence to analyze such datasets has become a current research focus in this domain. This review comprehensively elaborates on the fundamental principles of machine learning and its applications in the field of flexible wearable devices from five dimensions: supervised learning, unsupervised learning, deep learning, reinforcement learning, and meta-learning. Specifically, we provide an in-depth discussion of the significant research progress achieved in flexible wearable sensors using representative supervised learning algorithms, including Support Vector Machines (SVM), K-Nearest Neighbors (KNN), decision trees, random forests, Linear Discriminant Analysis (LDA), and Gaussian Naive Bayes. Additionally, we analyze the contributions of unsupervised learning methods such as Principal Component Analysis (PCA) and K-means clustering, as well as the wide-ranging applications of deep learning algorithms, including Deep Neural Networks (DNN), Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN) and Transformer model in flexible wearable sensing technology. Moreover, we highlight the potential applications of emerging machine learning techniques, such as meta-learning and reinforcement learning, in this domain. Finally, this review provides a thorough analysis and outlook on the challenges and opportunities faced by machine learning in the field of flexible wearable devices.