Knowledge Transfer with Weighted Adversarial Network for Cold-Start Store Site Recommendation

Store site recommendation aims to predict the value of store at candidate locations and then recommend optimal location company for placing a new brick-and-mortar store. Most existing studies focus on learning machine or deep models based large-scale training data chain stores in same city. However, expansion enterprises cities suffers from scarcity issues, these do not work city where no has been placed (i.e., cold-start problem). In this article, we propose unified approach recommendation, Weighted Adversarial Network with Transferability weighting scheme (WANT), transfer knowledge learned data-rich source target labeled data. particular, promote positive transfer, develop discriminator diminish distribution discrepancy between different distributions, which plays minimax game feature extractor learn transferable representations across by adversarial learning. addition, further reduce risk negative design transferability quantify examples reweight contribution relevant useful knowledge. We validate WANT using real-world dataset, experimental results demonstrate effectiveness our proposed model over several state-of-the-art baseline models.