Misfire Detection in Spark Ignition Engine Using Transfer Learning

Table 6

Performance of pretrained models for various learning rates.

Pre-trained model	Classification accuracy for different learning rate (%)			Overall accuracy (%)
	0.0001	0.0003	0.001
AlexNet	90.70	94.00	74.70	86.67
VGG-16	97.30	98.70	81.30	92.81
GoogLeNet	96.00	96.00	98.00	94.42
ResNet-50	97.60	96.00	94.40	95.44

Every pretrained network performs differently for the learning rates considered. The values highlighted in bold signify the best classification accuracies displayed by the pretrained networks for the variations in learning rate. The best performing network with highest overall classification accuracy is presented in the last column.