Table 1.
Experimental environment.
Fig 1.
Logistic regression algorithm positive and negative sample ratio on the test set AUC results graph.
Table 2.
Logistic regression algorithm positive and negative sample proportions on the test set accuracy results table.
Table 3.
Manual weighting results.
Fig 2.
Single model and fusion model results.
Table 4.
Fusion model for linear model construction AUC results table.
Fig 3.
Performance comparison of sales forecasting models in stable volatility mode.
Fig 4.
Comparison of prediction performance at different aggregation levels.
(a) High frequency customer set. (b) Low frequency customer set.
Fig 5.
Performance comparison of different models before and after feature selection.
Fig 6.
Comparison of different models of F1 oscillation curve.