Fig. 5: SVM classifications of high-humidity gases. | Microsystems & Nanoengineering

Fig. 5: SVM classifications of high-humidity gases.

From: Multiplexed DNA-functionalized graphene sensor with artificial intelligence-based discrimination performance for analyzing chemical vapor compositions

Fig. 5

a Reactivity for individual gas molecules of NH3, H2S, and NO at 2, 5, and 10 ppm and enhanced responses through FG-ssDNA under humid conditions. The bar with the dashes shows the reactivity in the case of 80% humidity, where an approximately 2 to 2.5 times higher reactivity is observed. Result of the subsequent b PCA and c result of SVM analysis for NH3-H2S mixed gases. The confusion matrix for d NH3-H2S, e NH3-NO, and f H2S-NO mixed gases after feature selection under conditions of high humidity. g Comparison of the classification rate of the Boruta algorithm with feature selection and SVM analysis without feature selection under conditions of high humidity

Back to article page