A Machine Learning-Based Model for Stability Prediction of Decentralized Power Grid Linked with Renewable Energy Resources

Table 6

Performance of ML models on oversample dataset.

Oversampling technique	Models	Accuracy (%)	Precision (%)	Recall (%)	-measure (%)	ROC (%)
ADASYN	ANN	95.4	97.5	93.4	95.4	99
	Averaged perceptron	79.2	78.6	81.3	79.9	87.7
	Bayes point machine	79.1	78.4	81.3	79.8	87.8
	Decision forest	91.2	93.7	88.8	91.2	97.2
	Decision jungle	89.5	89.7	89.7	89.7	96.4
	GBDT	95.9	96.5	95.4	96	99.4
	LightGBM	94.7	96.5	94	94.8	99.1
	Locally deep SVM	91.5	91.7	91.6	91.6	97.4
	LR	79.1	78.3	81.5	79.9	87.8
	SVM	78.8	78.5	80.6	79.5	87.6
	XGBoost	95.7	96.4	95	95.7	99.4
Borderline-SMOTE	ANN	95.1	95.8	94.4	95.1	99
	Averaged perceptron	76.9	77.5	75.9	76.7	86.1
	Bayes point machine	76.9	77.4	76.2	76.8	86.1
	Decision forest	90.8	93.5	87.6	90.5	96.8
	Decision jungle	89.3	91.2	87.1	89.1	96
	GBDT	95.4	96.9	93.7	95.3	99.3
	LightGBM	94.5	95.9	92.7	94.3	99
	Locally deep SVM	90.8	92.4	88.8	90.6	97.2
	LR	76.8	77.4	75.8	76.6	86.1
	SVM	76.8	77.6	75.6	76.6	85.9
	XGBoost	96.5	97.7	95.1	96.4	99.5
ROS	ANN	95.7	97.9	93.4	95.6	99.2
	Averaged perceptron	79.2	79.4	78.9	79.2	88.3
	Bayes point machine	79.4	79.7	79.2	79.4	88.4
	Decision forest	91.8	94.9	88.4	91.6	97.7
	Decision jungle	90.5	92	88.8	90.4	97
	GBDT	95.7	96.8	94.4	95.6	99.4
	LightGBM	95.7	97.2	93.9	95.5	99.4
	Locally deep SVM	90.8	92.4	88.8	90.6	97.2
	LR	79.4	79.5	79.3	79.4	88.4
	SVM	78.8	79.1	78.4	78.8	88.1
	XGBoost	96.8	97.5	95.9	96.7	99.6
SMOTE	ANN	95.6	97.8	93.4	95.5	99.3
	Averaged perceptron	80.1	80.7	79.3	80	89
	Bayes point machine	91.1	933	88.6	90.9	97.1
	Decision forest	95.9	97	94.7	95.8	99.3
	Decision jungle	90.4	90.3	90.7	90.5	96.5
	GBDT	80.3	80.9	79.5	80.2	89
	LightGBM	90	90.8	89.2	90	96.8
	Locally deep SVM	94.9	95.8	93.6	94.7	99.2
	LR	80.2	80.8	79.4	80.1	89
	SVM	79.8	80.6	78.5	79.6	88.9
	XGBoost	96.1	96.5	95.5	96	99.4