Skip to main content

Table 4 Deep neural network

From: Recognition of cooking activities through air quality sensor data for supporting food journaling

 

Original predictions

After post-processing

TN

29,3027

297,008

FP

41,140

37,159

TP

12,687

12,461

FN

3637

3863

Accuracy (%)

87.22

88.30

Specificity (%)

87.69

88.88

Recall (%)

77.72

76.34

Precision (%)

23.57

25.11

\(F_1\) score (%)

36.17

37.79

  1. Results with minute by minute modality, delayed recognition
  2. Numbers in italic indicate the largest value obtained in the experiment for each considered metric