Skip to main content
. 2020 Jun 23;20(12):3542. doi: 10.3390/s20123542

Figure 3.

Figure 3

Example of transformed feature characteristics extracted from the sensors by Principal Components Analysis (PCA). Features from all sensors are considered, and the best five features selected by the forward selection method are used as input to PCA. The four studied wine categories are marked by colors: LQ–low quality, AQ—average quality, HQ—high quality, and Ea—diluted ethanol.