Enhancing Photovoltaic Module Fault Diagnosis with Unmanned Aerial Vehicles and Deep Learning-Based Image Analysis

Table 11

Training (a), validation (b), and test (c) accuracies of lazy-based classifiers for various pretrained networks.

(a) Training accuracy of lazy-based classifiers for various pretrained networks

Classifier	Training accuracy (%)
	AlexNet	DenseNet-201	GoogleNet	ResNet-50	VGG16	VGG19
IBK	100.00	100.00	100.00	100.00	100.00	100.00
K-star	100.00	100.00	100.00	100.00	100.00	100.00
LWL	73.57	81.07	82.97	83.33	81.15	82.93

(b) Validation accuracy of lazy-based classifiers for various pretrained networks

Classifier	Validation accuracy (%)
	AlexNet	DenseNet-201	GoogleNet	ResNet-50	VGG16	VGG19
IBK	99.24	99.84	99.44	99.68	99.20	99.28
K-star	99.24	99.64	99.13	99.48	98.96	98.80
LWL	73.17	80.31	16.66	82.57	79.92	82.34

(c) Test accuracy of lazy-based classifiers for various pre-trained networks

Classifier	Test accuracy (%)
	AlexNet	DenseNet-201	GoogleNet	ResNet-50	VGG16	VGG19
IBK	99.84	100.00	99.80	99.68	99.52	99.84
K-star	99.52	100.00	99.04	99.52	99.52	99.36
LWL	76.03	80.00	82.85	82.22	77.77	80.63

Boldface entries represent the highest values obtained.