Skip to main content

Table 4 Classification accuracy of both single and poly microbial species in three different mediums using different classifier

From: In-vitro diagnosis of single and poly microbial species targeted for diabetic foot infection using e-nose technology

 

Blood agar (1,750 data)

Mueller Hinton (1,750 data)

Mac Conkey (1,000 data)

**Mixed media (4,500 data)

Classifiers

Train & validation (733 data)

Test (1017 data)

Train & validation (733 data)

Test (1017 data)

Train & validation (399 data)

Test (601 data)

Train & validation (1792 data)

Test (2708 data)

LDA

100%

100%

99.7%

98.7%

100%

100%

95.9%

94.7%

PNN

100%

100%

100%

89.7%

98.7%

98.2%

65.2%

63.3%

KNN

100%

100%

97.0%

95.4%

100%

100%

99.9%

99.2%

SVM

100%

100%

100%

100%

100%

100%

67.3%

65%

*PNN

100%

100%

100%

96.3%

100%

100%

99.2%

96.8%

*KNN

100%

100%

100%

100%

100%

100%

99.9%

99.6%

*SVM

100%

100%

100%

100%

100%

100%

98.9%

98.7%

  1. *Data was extracted from a discriminant function of LDA and subjected to classifier input.
  2. **Combination of three media (blood agar, Mueller Hinton & MacConkey), p value < 0.001.