Table 9 Quantitative analysis of K- fold cross validation of spectrofusionnet on ML classifiers.
From: SpectroFusionNet a CNN approach utilizing spectrogram fusion for electric guitar play recognition
Classifier | \(\:{F}_{late,concat}\) | \(\:{F}_{late,wavg}\) | \(\:{F}_{late,max}\) | |||
|---|---|---|---|---|---|---|
Mean Accuracy (%) | Mean F1-Score (%) | Mean Accuracy (%) | Mean F1-Score (%) | Mean Accuracy (%) | Mean F1-Score (%) | |
Random forest | 96.55 (96.55, 94.25, 95.40, 98.85, 97.70) | 95.12 (94.50, 93.20, 96.00, 97.80, 98.10) | 97.70 (98.85, 97.70, 98.85, 95.40, 97.70) | 97.68 (98.83, 97.69, 98.85, 95.36, 97.66) | 97.83 (100, 96.39, 96.39, 97.59, 98.78) | 97.83 (100, 95.15, 97.58, 97.61, 98.79) |
SVM | 95.10 (91.26, 92.40, 96.09, 97.14, 98.00) | 94.30 (90.40, 91.50, 95.00, 96.80, 97.80) | 96.09 (96.55, 96.55, 95.40, 95.40, 96.55) | 96.07 (96.53, 96.61, 95.38, 95.37, 96.49) | 97.34 (100, 93.98, 98.80, 96.39, 97.56) | 97.33 (100, 93.97, 98.80, 96.37, 97.52) |
KNN | 90.10 (86.67, 87.50, 90.80, 92.30, 93.24) | 89.20 (85.60, 86.80, 89.30, 91.10, 92.30) | 90.80 (95.40, 95.40, 85.06, 87.36, 90.80) | 90.77 (95.40, 95.42, 85.03, 87.28, 90.73) | 91.07 (98.80, 89.16, 81.93, 90.36, 95.12) | 91.01 (98.83, 89.16, 81.57, 90.59, 94.89) |
LMT | 96.92 (95.40, 96.00, 97.47, 98.20, 99.00) | 96.10 (94.80, 95.40, 96.70, 97.90, 98.50) | 97.47 (98.85, 100.00, 95.40, 95.40, 97.70) | 97.45 (98.83, 100.00, 95.38, 95.36, 97.66) | 98.79 (100, 97.59, 100, 98.80, 97.56) | 98.78 (100, 97.60, 100, 98.79, 97.52) |
Naive Bayes | 79.20 (79.08, 77.50, 78.39, 80.00, 81.20) | 77.80 (76.50, 75.60, 77.10, 78.90, 79.90) | 78.39 (77.01, 78.16, 74.71, 81.61, 80.46) | 78.32 (77.57, 77.84, 74.11, 81.52, 80.57) | 67.65 (73.49, 57.83, 63.86, 67.47, 75.61) | 67.42 (71.39, 58.54, 64.60, 66.69, 75.91) |
Linear SVM | 98.02 (96.55, 97.20, 98.16, 99.00, 99.04) | 97.40 (96.10, 97.00, 97.80, 98.50, 99.00) | 98.16 (98.85, 100.00, 97.70, 96.55, 97.70) | 98.15 (98.83, 100.00, 97.69, 96.55, 97.66) | 98.79 (100, 96.39, 100, 100, 97.56) | 98.79 (100, 96.36, 100, 100, 97.59) |
MLP | 97.50 (96.55, 96.80, 97.70, 98.50, 99.00) | 96.80 (96.10, 96.50, 97.20, 98.10, 98.90) | 97.70 (98.85, 100.00, 95.40, 96.55, 97.70) | 97.68 (98.83, 100.00, 95.37, 96.55, 97.66) | 97.34 (100, 93.98, 98.80, 97.59, 96.34) | 97.12 (98.83, 97.60, 96.44, 96.39, 96.35) |
Decision tree | 80.93 (78.39, 79.50, 80.00, 82.10, 82.56) | 79.80 (77.50, 78.20, 79.00, 81.00, 82.20) | 80.00 (83.91, 80.46, 80.46, 77.01, 78.16) | 79.91 (83.53, 80.85, 80.05, 76.81, 78.33) | 83.58 (79.52, 86.75, 87.95, 78.31, 85.37) | 81.60 (78.74, 83.81, 80.46, 80.70, 84.34) |
Gradient boosting | 92.60 (91.26, 90.50, 91.72, 93.40, 94.40) | 91.90 (90.00, 89.50, 91.30, 92.70, 93.80) | 91.72 (93.10, 93.10, 93.10, 89.66, 89.66) | 91.69 (92.96, 93.17, 93.00, 89.75, 89.54) | 90.35 (96.39, 83.13, 84.34, 91.57, 96.34) | 89.80 (96.31, 83.28, 84.22, 91.24, 93.95) |
AdaBoost | 74.88 (70.34, 72.50, 73.56, 76.20, 76.98) | 73.40 (69.80, 71.20, 72.50, 74.80, 75.90) | 73.56 (81.61, 75.86, 67.82, 62.07, 80.46) | 72.97 (81.90, 70.83, 69.68, 61.31, 81.13) | 75.37 (77.11, 74.70, 69.88, 75.90, 79.27) | 74.52 (77.28, 75.18, 67.51, 75.33, 77.34) |
