Stochastic constraints: How feasible is feasible?

Stochastic constraints: How feasible is feasible?. Eckman, D. J., Henderson, S. G., & Shashaani, S. In Corlu, C. G., Hunter, S. R., Lam, H., Onggo, B. S., Shortle, J., & Biller, B., editors, Proceedings of the 2023 Winter Simulation Conference, pages 3589-3600, Piscataway, New Jersey, 2023. IEEE.

Paper abstract bibtex 1 download

Stochastic constraints, which constrain an expectation in the context of simulation optimization, can be hard to conceptualize and harder still to assess. As with a deterministic constraint, a solution is considered either feasible or infeasible with respect to a stochastic constraint. This perspective belies the subjective nature of stochastic constraints, which often arise when attempting to avoid alternative optimization formulations with multiple objectives or an aggregate objective with weights. Moreover, a solution's feasibility with respect to a stochastic constraint cannot, in general, be ascertained based on only a finite number of simulation replications. We introduce different means of estimating how ``close'' the expected performance of a given solution is to being feasible with respect to one or more stochastic constraints. Some of these metrics entail a modest amount of numerical optimization. We explore how these metrics and their bootstrapped error estimates can be incorporated into plots showing a solver's progress over time when solving a stochastically constrained problem.

@inproceedings{eckhensha23,
	abstract = {Stochastic constraints, which constrain an expectation in the context of simulation optimization, can be hard to conceptualize and harder still to assess. As with a deterministic constraint, a solution is considered either feasible or infeasible with respect to a stochastic constraint. This perspective belies the subjective nature of stochastic constraints, which often arise when attempting to avoid alternative optimization formulations with multiple objectives or an aggregate objective with weights. Moreover, a solution's feasibility with respect to a stochastic constraint cannot, in general, be ascertained based on only a finite number of simulation replications. We introduce different means of estimating how ``close'' the expected performance of a given solution is to being feasible with respect to one or more stochastic constraints. Some of these metrics entail a modest amount of numerical optimization. We explore how these metrics and their bootstrapped error estimates can be incorporated into plots showing a solver's progress over time when solving a stochastically constrained problem.
},
	address = {Piscataway, New Jersey},
	author = {David J. Eckman and Shane G. Henderson and Sara Shashaani},
	booktitle = {Proceedings of the 2023 Winter Simulation Conference},
	date-added = {2016-01-10 16:07:54 +0000},
	date-modified = {2024-08-12 14:39:38 -0400},
	editor = {C. G. Corlu and S. R. Hunter and H. Lam and B. S. Onggo and J. Shortle and B. Biller},
	pages = {3589-3600},
	publisher = {IEEE},
	title = {Stochastic constraints: How feasible is feasible?},
	url_paper = {https://www.informs-sim.org/wsc23papers/302.pdf},
	year = {2023}}

Downloads: 1