摘要
Electronic nose (e-nose) is composed of a set of gas sensors combined with a series of algorithmic models. The practical application of the electronic nose system can prove that the electronic nose is more widely used in the classification problems, and always has a good performance. Moreover, it can be inferred that classification methods significantly influence e-nose. So far, the classification models proposed in e-nose can generally be divided into two categories. One is the linear classifier, representing the model of the Bayesian classifier, principal component analysis (PCA), and K-nearest neighbour (KNN), etc. The other is the nonlinear classifier, including support vector machine (SVM), random forest (RF), and extreme learning machine (ELM), etc. This review aims to supply a summary of the various classification methods used in e-nose, and provides a reference for the choice of an appropriate classification model used in e-nose in the specific application.
学科与关键词
文献加载失败,请重试
重试