Parsimonious modeling for sleep fidget soft sensor based on kernel methods

It is very normal that the soft sensor model based on computational intelligence would adopt so many secondary variables, so there is redundant correlative information in input space, it is easy to appear the curse of dimensionality, and the complexity of model is large. In order to obtain parsimonious sleep fidget soft sensor model that is high at precision and low at complexity, authors adopt the feature extraction methods such as principal component analysis (PCA) and kernel principal component analysis (KPCA) to decrease the input dimensionalities of the model, and build five kinds of model for sleep fidget soft sensor. Simulation results show that these models all have high precision, and it is very easy to select a simple and effective model to implement by hardware. Therefore, the method can not only overcome the correlativity among inputs variables, decrease the dimensionality of the model, but also reduce the complexity of the model, promote the soft sensor model to implement by hardware.