Robust sensitivity analysis for stochastic systems
From MaRDI portal
Abstract: We study a worst-case approach to measure the sensitivity to model misspecification in the performance analysis of stochastic systems. The situation of interest is when only minimal parametric information is available on the form of the true model. Under this setting, we post optimization programs that compute the worst-case performance measures, subject to constraints on the amount of model misspecification measured by Kullback-Leibler (KL) divergence. Our main contribution is the development of infinitesimal approximations for these programs, resulting in asymptotic expansions of their optimal values in terms of the divergence. The coefficients of these expansions can be computed via simulation, and are mathematically derived from the representation of the worst-case models as changes of measure that satisfy a well-defined class of functional fixed point equations.
Let \(X_1,\dots,X_T\) be a sequence of indipendent identically distributed random variables, \(h(.)\) be a real-valued function and \(T\) the time horizon. Let \(h\) be evaluated given to its arguments, but it does not necessarily have the closed form. Moreover, assume a baseline model \(P_0\) approximately describing each \(X_i\), so that the corresponding baseline performance measure is \(E_0\,h(X_T)\). On the other hand, let \(P_f\) denote the distribution that governs the true model, which is unknown, and \(E_f\) the orresponding expectation. The author is interested in the worst (best) case optimization problem NEWLINE\[NEWLINE\max/\min\;E_f\,h(X_T) \quad st\quad D(P_f||P_0)\leq\eta,\;NEWLINE\]NEWLINE where \(D(P_f||P_0)\) denotes the Kullback-Leibler divergence. The main result of this paper stipulates that, when letting \(\eta\to0\), then under mild technical assumptions on~\(h\) the optimal values of the problem can be written in the form NEWLINE\[NEWLINE \max/\min\;E_f\,h(X_T)=E_0\,h(X_T) + \zeta_1(P_0,h)\sqrt{\eta}+\zeta_2(P_0,h)\eta+\dots NEWLINE\]NEWLINE where \(\zeta_1(P_0,h),\,\zeta_2(P_0,h)\) is a sequence of coefficients that can be written explicitly in terms of~\(h\) and~\(P_0\).
Recommendations
- Robust analysis in stochastic simulation: computation and performance guarantees
- Generalized sensitivity analysis of ergodic stochastic systems
- Stochastic Control with Imperfect Models
- Estimates of performance sensitivity of a stochastic system
- Generalized estimates for performance sensitivities of stochastic systems
Cites work
- A new look at the statistical model identification
- A Perturbation Theory for Ergodic Markov Chains and Application to Numerical Approximations
- Distributionally robust optimization and its tractable approximations
- Distributionally robust optimization under moment uncertainty with application to data-driven problems
- Efficient nested simulation for estimating the variance of a conditional expectation
- Estimation of Entropy and Mutual Information
- Global sensitivity analysis: The primer
- scientific article; zbMATH DE number 1092005 (Why is no real title available?)
- Minimax optimal control of stochastic uncertain systems with relative entropy constraints
- On Information and Sufficiency
- On the optimality of threshold control in queues with model uncertainty
- On the Optimum Rate of Transmitting Information
- Performance analysis approximation in a queueing system of type \(M/G/1\)
- Perturbation theory and finite Markov chains
- Probabilistic Sensitivity Analysis of Complex Models: A Bayesian Approach
- Relative Entropy, Exponential Utility, and Robust Dynamic Pricing
- Robust Control of Markov Decision Processes with Uncertain Transition Matrices
- Robust Dynamic Programming
- Robust portfolio control with stochastic factor dynamics
- Robust risk measurement and model risk
- Sensitivity and convergence of uniformly ergodic Markov chains
- Series expansions for continuous-time Markov processes
- SERIES EXPANSIONS FOR FINITE-STATE MARKOV CHAINS
- Taylor series expansions for stationary Markov chains
- The Influence Curve and Its Role in Robust Estimation
- The Maclaurin series for performance functions of Markov chains
- Tractable stochastic analysis in high dimensions via robust optimization
- Unbiased estimation with square root convergence for SDE models
- Variance based sensitivity analysis of model output. Design and estimator for the total sensitivity index
Cited in
(61)- Work by Robert Kalaba on automated sensitivity analysis
- Stochastic sensitivity analysis
- Rigorous sensitivity analysis for parameter-dependent systems of equations
- The impact of model risk on dynamic portfolio selection under multi-period mean-standard-deviation criterion
- Scalable information inequalities for uncertainty quantification
- Asymptotic optimality of the generalized \(c\mu\) rule under model uncertainty
- Toward theoretical understandings of robust Markov decision processes: sample complexity and asymptotics
- Frameworks and results in distributionally robust optimization
- Model risk in mean-variance portfolio selection: an analytic solution to the worst-case approach
- Robust quantile estimation under bivariate extreme value models
- System design of stochastic models using robustness of temporal properties
- Robust empirical optimization is almost the same as mean-variance optimization
- Estimates of performance sensitivity of a stochastic system
- Variance regularization in sequential Bayesian optimization
- Sensitivity analysis for systems with structured uncertainty
- Reducing parameter uncertainty for stochastic systems
- Distributionally robust optimization with principal component analysis
- A Bayesian risk approach to data-driven stochastic optimization: formulations and asymptotics
- Decentralized sensitivity of stochastic composite systems
- Statistics of robust optimization: a generalized empirical likelihood approach
- Uncertainty quantification for Markov processes via variational principles and functional inequalities
- Robust analysis in stochastic simulation: computation and performance guarantees
- Quantification of model uncertainty on path-space via goal-oriented relative entropy
- Formulation and properties of a divergence used to compare probability measures without absolute continuity
- Calibration of distributionally robust empirical optimization models
- (f,)-divergences: interpolating between f-divergences and integral probability metrics
- Optimal ergodic harvesting under ambiguity
- Confidence regions of stochastic variational inequalities: error bound approach
- Subsampling to Enhance Efficiency in Input Uncertainty Quantification
- Robust Simulation with Likelihood-Ratio Constrained Input Uncertainty
- Robustness to incorrect system models in stochastic control
- Asymptotic analysis of a multiclass queueing control problem under heavy traffic with model uncertainty
- Randomized Global Sensitivity Analysis and Model Robustness
- Recovering best statistical guarantees via the empirical divergence-based distributionally robust optimization
- Optimization-based calibration of simulation input models
- Quantifying distributional model risk via optimal transport
- Brownian control problems for a multiclass M/M/1 queueing problem with model uncertainty
- Robustness to Incorrect Priors in Partially Observed Stochastic Control
- scientific article; zbMATH DE number 6286911 (Why is no real title available?)
- Robust Actuarial Risk Analysis
- Distributionally robust portfolio maximization and marginal utility pricing in one period financial markets
- Sensitivity of Multiperiod Optimization Problems with Respect to the Adapted Wasserstein Distance
- Optimal Dividends Under Model Uncertainty
- Minimal Kullback–Leibler Divergence for Constrained Lévy–Itô Processes
- Data-driven aerodynamic shape design with distributionally robust optimization approaches
- Parametric scenario optimization under limited data: a distributionally robust optimization view
- Residuals-based distributionally robust optimization with covariate information
- Robust distortion risk measures
- Sensitivity of functionals of McKean-Vlasov SDEs with respect to the initial distribution
- On the sensitivity of restless bandits solutions to uncertainty in the models of the arms
- Distributionally robust optimization and robust statistics
- Distributionally robust optimization
- On regularization schemes for data-driven optimization based on Cressie–Read divergence and CVaR
- Sensitivity analysis of Wasserstein distributionally robust optimization problems
- Sensitivity analysis on policy-augmented graphical hybrid models with Shapley value estimation
- Optimal investment and equilibrium pricing under ambiguity
- The distributionally robust prediction error of the \(\sqrt{\mathrm{LASSO}}\) and related estimators
- Calibrating non-identifiable high-dimensional simulation models: a framework via eligibility set
- Distributionally robust risk evaluation with an isotonic constraint
- Learning and decision-making with data : optimal formulations and phase transitions
- Learning models with uniform performance via distributionally robust optimization
This page was built for publication: Robust sensitivity analysis for stochastic systems
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2833103)
