PREFER: Point-of-interest REcommendation with efficiency and privacy-preservation via Federated ...
summary
Proposes PREFER, an edge-accelerated federated learning system that keeps check-ins local, shares non-sensitive multi-dimensional parameters and aggregates updates on nearby edge servers (e.g., base stations) rather than in the cloud. Implements a prototype and tests two POI models on two real-world datasets using Hit Ratio and NDCG. Reports near-centralized recommendation quality with only a small drop versus cloud training and substantially reduced end-to-end time and latency compared to cloud aggregation. improves efficiency and privacy-preservation
AI-written summary · about
related talks
06mLearning from Hometown and Current City: Cross-city POI Recommendation via Interest Drift and ...
06mPMF: A Privacy-preserving Human Mobility Prediction Framework via Federated Learning
48mFederated ML for Edge Applications
07mDecentralized Attention-based Personalized Human Mobility Prediction
34mTraining without sharing: Is Federated Learning the answer to data privacy in AI?
51mExploiting Similarity in Federated Learning
28mA Privacy Enhancing Federated Learning Framework for Multi-Party Distributed ..
27mCreate Value and Insight Through Privacy-preserving Data Collaboration [...]