“…Support vector machines [ 17 , 21 , 30 ] and random forest (RF) [ 1 , 14 , 16 ] were the second most used models. Furthermore, algorithms such as K-means [ 30 ], multimodal regression [ 14 ], boosting [ 14 ], dynamic time warping [ 13 ], k-nearest neighbors [ 20 ], sliced inverse regression (SIR) [ 20 ], decision tree [ 23 ], Naive Bayes [ 21 ], SimpleLogistic [ 21 ], linear discriminant analysis [ 20 ], principal component analysis (PCA) [ 20 ], and triangle center [ 24 ] were deployed in the papers to classify sitting postures. In this study, 6 papers [ 14 , 15 , 18 , 20 , 21 , 30 ] compared the performance of classification algorithms using more than one classifier.…”