Get Started
Home
Topics
Search
Library
Research questionHow can human mobility predictability be estimated and compared when its ground truth is unobservable?The best achievable prediction performance cannot be observed directly in real mobility data, leaving no definitive reference for estimator accuracy. Comparisons are further complicated when mobility is represented in fundamentally different forms.
AI
Evaluation & Benchmarks
Machine Learning
Research Paper
Statistical Machine Learning
Technology
Latest papersRecent research connected to this question, newest first.BER-PEF: Unified Human Mobility Predictability Evaluation via Bayes Error Rate EstimationThe source presents BER-PEF, a Bayes-error-rate-based evaluation framework that maps heterogeneous representations into a shared feature–label space and compares estimates using controlled perturbations and a common predictability reference interval. Evidence comes from Foursquare NYC and TKY, GeoLife, and T-Drive data; it reports reference-discrepancy results for symbolic sequences and numeric trajectories, and examines contextual inputs and learned representations under the same protocol.research paper · Sep 3, 2026
Related questions
How can urban mobility demand be forecast across modes with mismatched spatial structures and scarce target histories?How can we tell whether cross-layer routing predictability reflects shared routing dynamics or generic hidden-state smoothness?How can human motion be retargeted across different humanoid morphologies without hand-crafted mappings or losing pose detail?How can wearable-derived mobility outcomes be modelled longitudinally across diseases without paired participants?