Related Experiment Video
Updated: May 28, 2025
11:53
Spatial Multiobjective Optimization of Agricultural Conservation Practices using a SWAT Model and an Evolutionary Algorithm
Published on: December 9, 2012
12.9K
Optimization design of cross border intelligent marketing management model based on multi layer perceptron-grey wolf
Zongping Lin1, Jing Yang2, Yabin Lian3
1School of Economics and Management, Quanzhou University of Information Engineering, Quanzhou, 362008, Fujian, China.
Scientific Reports
|February 11, 2025
Summary
This study introduces a novel cross-border intelligent marketing model, MLP-GWO-CNN, enhancing personalized recommendations by integrating user ratings and labels. The model significantly improves marketing accuracy and personalization for e-commerce.
Area of Science:
- Artificial Intelligence
- Machine Learning
- E-commerce Analytics
Background:
- Traditional linear models struggle with implicit information in complex cross-border marketing scenarios, limiting personalization.
- Existing methods often fail to effectively extract high-order features from sparse and complex user/market data.
Purpose of the Study:
- To develop an advanced cross-border intelligent marketing model for improved personalization and accuracy.
- To address the limitations of traditional models in handling implicit information and complex data.
Main Methods:
- Proposed a dual-path deep network: Multi-layer Perceptron (MLP) for user interest features and Convolutional Neural Networks (CNN) for semantic label features.
- Integrated Grey Wolf Optimization (GWO) to optimize the MLP, mitigating sensitivity to initial values and local optima.
- Fused latent feature vectors from MLP and CNN for final predictive marketing strategy generation.
Main Results:
- The MLP-GWO-CNN model demonstrated superior performance over traditional recommendation algorithms on a real cross-border e-commerce dataset.
- Achieved over 89% accuracy and 90% recall rate, indicating effective utilization of user tag information.
- Significantly improved the accuracy and personalization of marketing recommendations.
Conclusions:
- The proposed MLP-GWO-CNN model effectively handles complex, sparse marketing data by extracting implicit high-order information.
- The integration of MLP, GWO, and CNN offers a robust solution for personalized cross-border intelligent marketing.
- The model shows significant potential for enhancing e-commerce marketing strategies through advanced machine learning techniques.
