Table 1 Summary of the results with two datasets
From: CRNet: a multimodal deep convolutional neural network for customer revisit prediction
Models | Class | Precision | Recall | F1-Score | Accuracy |
|---|---|---|---|---|---|
Dataset: CRD | |||||
Text-image fusion model | Revisit | 0.8753 | 0.7825 | 0.8263 | 0.7417 |
Not | 0.4266 | 0.5921 | 0.4959 | ||
MVAN | Revisit | 0.9280 | 0.9478 | 0.9378 | 0.9012 |
Not | 0.7926 | 0.7308 | 0.7605 | ||
CRNet | Revisit | 0.9700 | 0.9760 | 0.9730 | 0.9575 |
Not | 0.9102 | 0.8896 | 0.8998 | ||
Dataset: MFDRD | |||||
Text-image fusion model | Revisit | 0.7725 | 0.7215 | 0.7461 | 0.7065 |
Not | 0.6231 | 0.6843 | 0.6522 | ||
MVAN | Revisit | 0.9364 | 0.9106 | 0.9098 | 0.8951 |
Not | 0.8415 | 0.8846 | 0.8747 | ||
CRNet | Revisit | 0.9915 | 0.9136 | 0.9509 | 0.9436 |
Not | 0.8850 | 0.9883 | 0.9338 | ||