Detecting and Identifying Industrial Gases by a Method Based on Olfactory Machine at Different Concentrations

Figure 3

Classification results of steady-state PCA, steady-state LDA, and steady-state PCA + LDA method before the feature extraction corresponding to (a), (b), and (c), respectively. The horizontal axis represents the proportion of the first variance (eigenvalues corresponding to the eigenvalues), and the vertical axis represents the proportion of the second variance, and the two constitute the original classification in the two-dimensional distribution space.

(a)

(b)

(c)