Optimization-based calibration of simulation input models

DOI10.1287/OPRE.2018.1801zbMATH Open1444.90027arXiv1605.09074OpenAlexW2964745183MaRDI QIDQ5129200FDOQ5129200

Authors: Aleksandrina Goeva, Henry Lam, Huajie Qian, Bo Zhang 

Publication date: 26 October 2020

Published in: Operations Research (Search for Journal in Brave)

Abstract: Studies on simulation input uncertainty often built on the availability of input data. In this paper, we investigate an inverse problem where, given only the availability of output data, we nonparametrically calibrate the input models and other related performance measures of interest. We propose an optimization-based framework to compute statistically valid bounds on input quantities. The framework utilizes constraints that connect the statistical information of the real-world outputs with the input-output relation via a simulable map. We analyze the statistical guarantees of this approach from the view of data-driven robust optimization, and show how the guarantees relate to the function complexity of the constraints arising in our framework. We investigate an iterative procedure based on a stochastic quadratic penalty method to approximately solve the resulting optimization. We conduct numerical experiments to demonstrate our performance in bounding the input models and related quantities.

Full work available at URL: https://arxiv.org/abs/1605.09074

Recommendations

• Input-output uncertainty comparisons for discrete optimization via simulation
• Reducing Simulation Input-Model Risk via Input Model Averaging
• A Bayesian framework for quantifying uncertainty in stochastic simulation
• Robust analysis in stochastic simulation: computation and performance guarantees
• Quantifying input uncertainty via simulation confidence intervals

zbMATH Keywords

uncertainty quantificationmodel calibrationdistributionally robust optimizationinput modeling

Mathematics Subject Classification ID

Sensitivity, stability, parametric optimization (90C31) Transportation, logistics and supply chain management (90B06) Mathematical modeling or simulation for problems pertaining to operations research and mathematical programming (90-10)

Cites Work

• Weak convergence and empirical processes. With applications to statistics
• Bayesian calibration of computer models. (With discussion)
• Mirror descent and nonlinear projected subgradient methods for convex optimization.
• Robust Stochastic Approximation Approach to Stochastic Programming
• Title not available (Why is that?)
• Why least squares and maximum entropy? An axiomatic approach to inference for linear inverse problems
• Theory and applications of robust optimization
• Title not available (Why is that?)
• Robust optimization
• Relative Entropy, Exponential Utility, and Robust Dynamic Pricing
• Accelerated gradient methods for nonconvex nonlinear and stochastic programming
• Stochastic block mirror descent methods for nonsmooth and stochastic optimization
• Distributionally robust optimization under moment uncertainty with application to data-driven problems
• Title not available (Why is that?)
• Generalized Chebychev Inequalities: Theory and Applications in Decision Analysis
• Optimal Inequalities in Probability Theory: A Convex Optimization Approach
• Stochastic First- and Zeroth-Order Methods for Nonconvex Stochastic Programming
• A Semidefinite Programming Approach to Optimal-Moment Bounds for Convex Classes of Distributions
• Mini-batch stochastic approximation methods for nonconvex stochastic composite optimization
• Distributionally robust optimization and its tractable approximations
• Estimation for an M/G/  queue with incomplete information
• Inverse Problem Theory and Methods for Model Parameter Estimation
• Robust Dynamic Programming
• Robustness
• General bounds and finite-time improvement for the Kiefer-Wolfowitz stochastic approximation algorithm
• Multidimensional Stochastic Approximation Methods
• Title not available (Why is that?)
• Stochastic optimisation with inequality constraints using simultaneous perturbations and penalty functions
• Robust portfolio control with stochastic factor dynamics
• Probabilistic inversion of expert judgments in the quantification of model uncertainty
• Fitting birth-and-death queueing models to data
• Limit theorems for \(U\)-processes
• A semidefinite optimization approach to the steady-state analysis of queueing systems
• Approximating a Point Process by a Renewal Process, I: Two Basic Methods
• Estimating characteristics of queueing networks using transactional data
• Maximum likelihood estimation for single server queues from waiting time data
• Estimation for queues from queue length data
• Verification and validation of simulation models
• Non-parametric estimation for the \(M/G/\infty\) queue
• Nonparametric Inference About Service Time Distribution from Indirect Measurements
• WEIGHTED MONTE CARLO: A NEW TECHNIQUE FOR CALIBRATING ASSET-PRICING MODELS
• Distributionally robust Markov decision processes
• On the optimality of threshold control in queues with model uncertainty
• Approximating a Point Process by a Renewal Process: The View Through a Queue, an Indirect Approach
• Moment estimation of customer loss rates from transactional data
• The Queue Inference Engine: Deducing Queue Statistics from Transactional Data
• Queue inference from periodic reporting data
• The Ocean Circulation Inverse Problem
• Robust risk measurement and model risk
• Distributionally Robust Convex Optimization
• Minimax optimal control of stochastic uncertain systems with relative entropy constraints
• Large Sample Properties of Weighted Monte Carlo Estimators
• Tail analysis without parametric models: a worst-case perspective
• Ambiguous joint chance constraints under mean and dispersion information
• The empirical likelihood approach to quantifying uncertainty in sample average approximation
• Robust sensitivity analysis for stochastic systems
• Robust analysis in stochastic simulation: computation and performance guarantees
• Quantifying distributional model risk via optimal transport
• Maximum likelihood estimates and confidence intervals of an \(M/M/R\) queue with heterogeneous servers
• Parameter estimation using partial information with applications to queueing and related models
• Ambiguous risk constraints with moment and unimodality information
• Two-point methods for assessing variability in simulation output
• Calculation of confidence intervals for simulation output
• Recovering best statistical guarantees via the empirical divergence-based distributionally robust optimization
• Robust sample average approximation
• Bounding wrong-way risk in CVA calculation
• Quantifying input uncertainty via simulation confidence intervals
• Nonparametric inference of photon energy distribution from indirect measurement

Cited In (9)

• Bayesian sequential data collection for stochastic simulation calibration
• Input optimization: I. Optimal realizations of mathematical models
• Robust analysis in stochastic simulation: computation and performance guarantees
• Maximum likelihood estimation by Monte Carlo simulation: toward data-driven stochastic modeling
• Input-output uncertainty comparisons for discrete optimization via simulation
• Constrained multifidelity optimization using model calibration
• Bayesian optimisation vs. input uncertainty reduction
• Globally optimized calibration of environmental models
• Efficient calibration for imperfect computer models

Uses Software

• ROME

This page was built for publication: Optimization-based calibration of simulation input models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5129200)