Monte carlo algorithms for expected utility estimation in dynamic purchasing
Bibliographic record
Abstract
This thesis describes a theory for decision-making in a dynamic purchasing environment where one of possibly many bundles of items must be purchased from possibly many suppliers. The online combinatorial purchasing problem is defined as the problem with which a purchase agent in such an environment is faced when deciding which combination of items, from whom and at what time to buy in order to maximize overall satisfaction. Expected utility maximization is used as the criterion on which decision-making is based. To facilitate the exchange of probabilistic and temporal information between suppliers and purchasers, the PQR protocol is defined. The theory considers two distinct purchasing models, one in which only complete bundle purchases can be made at any time, and one in which partial bundle purchases are allowed. In the first model, a decision procedure that exploits future time intervals where several options will be available is developed that provably yields higher expected utility than simply pursuing the bundle with highest expected utility. For the second model, the QR-tree is defined as a decision tree that can be exponentially smaller than conventional decision trees when used to model the same system of decisions. Efficient Monte Carlo methods are developed that solve the QR-tree in linear time on the number of nodes in the tree. Results show that these methods estimate expected utility as much as 25 times more accurately than a greedy method that always pursues the bundle with the current highest expected utility.
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 | 0.008 | 0.033 |
| Meta-epidemiology (narrow) | 0.001 | 0.002 |
| Meta-epidemiology (broad) | 0.003 | 0.001 |
| Bibliometrics | 0.002 | 0.003 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.002 | 0.004 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.002 | 0.004 |
| Insufficient payload (model declined to judge) | 0.006 | 0.001 |
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".