Table 1.
Key theories in customer loyalty.
Table 2.
Literature review.
Table 3.
Prediction models.
Table 4.
Some studies related to customer behaviors using learning techniques.
Table 5.
Model approaches.
Table 6.
Process of addressing correlation and determining feature weights using SEM.
Fig 1.
Overall view of the steps performed.
Table 7.
Variables used for learning techniques.
Table 8.
Summary of customer information.
Table 9.
Development and detailed explanations for each variable.
Table 10.
Hyperparameter settings.
Table 11.
Evaluation criteria.
Table 12.
Correlation matrix of customer features.
Table 13.
Weights of composite variables for loyalty.
Table 14.
Different methods for predicting customer loyalty.
Table 15.
Different methods for predicting customer loyalty by neighborhood.
Fig 2.
Comparison of model performance for predicting customer loyalty based on neighborhood.
Fig 3.
Comparison of model performance for predicting customer loyalty.
Model 1: Classification Model and Autoencoder. Model 2: CNN. Model 3: Autoencoder. Model 4: PCA & Random Forest. Model 5: XGBoost & LSTM. Model 6: PSO and LSTM. Model 7: Adaptive GCN-Transformer Hybrid.
Fig 4.
Evaluation of models for predicting neighborhood loyalty with ROC curve.
Model 1: XGBoost & LSTM. Model 2: PSO and LSTM. Model 3: PCA & Random Forest. Model 4: CNN.
Fig 5.
Evaluation of models for predicting individual customer loyalty with ROC curve.
Model 1: PCA & Random Forest. Model 2: CNN. Model 3: Classification Model and Autoencoder. Model 4: XGBoost & LSTM. Model 7: Adaptive GCN-Transformer Hybrid.
Fig 6.
Comparison of neighborhood loyalty prediction models with confusion matrix.
Fig 7.
Comparison of individual customer loyalty prediction models with confusion matrix.
Fig 8.
Correlation plot of features with loyalty prediction.
Table 16.
Interpretation of correlations between variables and loyalty.
Table 17.
Performance analysis of the recommendation system.
Table 18.
Evaluation criteria of the baseline model.
Table 19.
Impact scenarios on key factors.
Fig 9.
Comparison of the baseline model and advanced models in predicting customer loyalty.
Fig 10.
Comparison of the baseline model and advanced models in predicting neighborhood loyalty.
Model 1: Baseline Model. Model 2: Classification Model and Autoencoder. Model 3: CNN. Model 4: Autoencoder. Model 5: PCA & Random Forest. Model 6: XGBoost & LSTM. Model 7: PSO and LSTM. Model 8: Adaptive GCN-Transformer Hybrid.
Fig 11.
Effects of different scenarios on loyalty.
Fig 12.
Simulation of different scenarios and their impact on customer loyalty.
Fig 13.
Architecture of the loyalty prediction model integrated into an Enterprise Information System (EIS).