Taming Reasoning in Temporal Probabilistic Relational Models
This work addresses a computational bottleneck for researchers and practitioners using temporal probabilistic relational models, offering an incremental improvement in inference efficiency.
The paper tackles the problem of reasoning becoming infeasible in temporal probabilistic relational models due to evidence grounding over time, and presents temporal approximate merging (TAMe) to restore a lifted representation, resulting in significantly improved runtime performance with small, bounded errors.
Evidence often grounds temporal probabilistic relational models over time, which makes reasoning infeasible. To counteract groundings over time and to keep reasoning polynomial by restoring a lifted representation, we present temporal approximate merging (TAMe), which incorporates (i) clustering for grouping submodels as well as (ii) statistical significance checks to test the fitness of the clustering outcome. In exchange for faster runtimes, TAMe introduces a bounded error that becomes negligible over time. Empirical results show that TAMe significantly improves the runtime performance of inference, while keeping errors small.