Prediction of probable backorder scenarios in the supply chain using Distributed Random Forest and Gradient Boosting Machine learning techniques

Journal of Big Data

Table 6 Models’ characteristics during the testing phase

Performance metric	GBM trained with ranged data	GBM trained with actual data	DRF trained with ranged data	DRF trained with actual data
Classification accuracy	0.9892	0.7919	0.9835	0.8436
Precision	0.8869	0.6896	0.8231	0.7213
Recall/sensitivity	0.7849	0.5876	0.8488	0.6893
Specificity	0.9964	0.7991	0.9986	0.8407
F1 score	0.7845	0.6345	0.8357	0.7049
Misclassification error	0.0107	0.2080	0.0164	0.1563