Self-Adapting Network Relaxations for Weakly Coupled Markov Decision Processes
Bibliographic record
Abstract
High-dimensional weakly coupled Markov decision processes (WDPs) arise in dynamic decision making and reinforcement learning, decomposing into smaller Markov decision processes (MDPs) when linking constraints are relaxed. The Lagrangian relaxation of WDPs (LAG) exploits this property to compute policies and (optimistic) bounds efficiently; however, dualizing linking constraints averages away combinatorial information. We introduce feasibility network relaxations (FNRs), a new class of linear programming relaxations that exactly represents the linking constraints. We develop a procedure to obtain the unique minimally sized relaxation, which we refer to as self-adapting FNR, as its size automatically adjusts to the structure of the linking constraints. Our analysis informs model selection: (i) the self-adapting FNR provides (weakly) stronger bounds than LAG, is polynomially sized when linking constraints admit a tractable network representation, and can even be smaller than LAG, and (ii) self-adapting FNR provides bounds and policies that match the approximate linear programming (ALP) approach but is substantially smaller in size than the ALP formulation and a recent alternative Lagrangian that is equivalent to ALP. We perform numerical experiments on constrained dynamic assortment and preemptive maintenance applications. Our results show that self-adapting FNR significantly improves upon LAG in terms of policy performance and/or bounds, while being an order of magnitude faster than an alternative Lagrangian and ALP, which are unsolvable in several instances. This paper was accepted by Baris Ata, stochastic models and simulation. Supplemental Material: The electronic companion and data files are available at https://doi.org/10.1287/mnsc.2022.01108 .
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.012 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.003 |
| Insufficient payload (model declined to judge) | 0.007 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".