Skip to main content
. 2021 Oct 8:1–11. Online ahead of print. doi: 10.1007/s13369-021-06041-4

Table 3.

Results achieved on the testing set for the vowels /a/, /e/ and /o/

Algorithm Sensibility (%) Specificity (%) Accuracy (%) Precision (%) F1-score (%) AUC
Naive Bayes [25] 100.00 35.29 67.65 60.71 75.56 0.706
Bayes Net [25] 88.24 29.41 58.82 55.56 68.18 0.671
SVM [6] 94.12 52.94 73.53 66.67 78.05 0.735
SGD [3] 94.12 47.06 70.59 64.00 76.19 0.706
Ibk [2] 88.24 47.06 67.65 62.50 73.17 0.676
LWL [14] 58.82 52.94 55.88 55.56 57.14 0.647
Adaboost [12] 70.59 76.47 73.53 75.00 72.73 0.785
Bagging [12] 76.47 64.71 70.59 68.42 72.22 0.747
OneR [19] 29.41 76.47 52.94 55.56 38.46 0.529
Decision Table [26] 64.71 47.06 55.88 55.00 59.46 0.554
J48 [41] 64.71 52.94 58.82 57.89 61.11 0.588
Random Forest [49] 94.12 70.59 82.35 76.19 84.21 0.901