Optimizing Production Decisions Using a Hybrid Simulation–Genetic Algorithm Approach
Bibliographic record
Abstract
Mathematical programming has for a long time been recognized as a powerful tool. Despite its capacity for solving constrained optimization problems under uncertainty, some methodological obstacles have persisted over the years. The main problem is that the eventually complex results of an unbiased statistical analysis (multiple correlated stochastic variables with different distributions and nonadditive links between) cannot be adequately accounted for within minimization of total absolute deviation (MOTAD) or expected value‐variance (EV) models that rely on the algorithmic determination of the variability measure. In this paper, we develop a methodological hybrid consisting of Monte Carlo simulation and genetic algorithms: the Monte Carlo simulation facilitates the easy representation of diverse stochastic processes and correlation, and the genetic algorithm ensures that the optimization procedure remains applicable even in the case of complex stochastic information. This hybrid approach is applied to the production‐planning problem of a German crop farm. Variant calculations are used to account for the unknown risk attitude of the farmer. Model results demonstrate that optimized production programs and expected total gross margins are not only highly sensitive to the risk attitude, but also to the stochastic processes that are estimated (or assumed) for various activities. We furthermore find evidence that the hybrid approach is able to generate considerable improvement in farm‐program decisions and outperforms planning models that assume static distributions. La programmation mathématique est reconnue depuis longtemps comme étant un outil puissant. Malgré sa capacitéà résoudre des problèmes d'optimisation avec contraintes en situation d'incertitude, certains obstacles méthodologiques ont persisté au fil du temps. Le principal problème réside dans le fait que les résultats éventuellement complexes d'une analyse statistique non biaisée (plusieurs variables aléatoires corrélées avec différentes distributions et des liens non additifs entre elles) ne peuvent être adéquatement représentés dans les modèles MOTAD et E‐V qui dépendent de la détermination algorithmique de la mesure de la variabilité. Dans le présent article, nous avons élaboré une méthode hybride à partir d'une simulation de Monte Carlo et d'algorithmes génétiques: la simulation de Monte Carlo facilite la représentation de divers processus stochastiques et de diverses corrélations, et l'algorithme génétique assure que la procédure d'optimisation demeure applicable même dans le cas d'information stochastique complexe. Cette méthode hybride est appliquée au problème de planification de la production auqúel est confrontée une exploitation de cultures en Allemagne. Des calculs de variantes sont utilisés pour tenir compte de l'attitude inconnue du producteur quant au risque. Les résultats du modèle indiquent que les programmes de production optimisés et les marges brutes totales prévues ne sont pas uniquement sensibles à l'attitude face aux risques mais aussi aux processus stochastiques qui sont estimés (ou supposés) pour diverses activités. Nous avons également trouvé que la méthode hybride peut améliorer considérablement les décisions concernant les programmes agricoles et qu'elle est supérieure aux modèles de planification qui supposent des distributions statiques.
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 | 0.001 | 0.002 |
| 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 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.002 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".