Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 20 of 78 for “"Distributionally robust optimization"”.
-
Distributionally Robust Optimization for Reinforcement Learning
… In this dissertation, our goal is to utilize distributionally robust optimization (DRO) to overcome the above-mentioned limitations of RL, and to develop novel and practical RL algorithms with improved robustness and performance. To achieve the goal, we follow two main objectives. The first …
-
DISTRIBUTIONALLY ROBUST OPTIMIZATION WITH INFINITELY CONSTRAINED AMBIGUITY SETS
… description. We study both static and adaptive distributionally robust optimization models with infinitely constrained ambiguity sets. We propose a tractable solution procedure to obtain approximate solutions, which can be iteratively refined. We present a unified and tractable framework for …
-
Regularized distributionally robust optimization with applications in finance
… resulting in large prediction errors. Distributionally robust optimization (DRO) is a successful candidate for dealing with distributional uncertainty. It considers the optimization of the worst-case scenario in an ambiguity set of probability distributions whose size is quantified by …
-
Adapted Wasserstein distances and applications to distributionally robust optimization
… Wasserstein distances and their applications to distributionally robust optimization (DRO) problems in a dynamic context. In Chapter 3, we propose a transfer principle to study the adapted 2-Wasserstein distance between stochastic processes. We obtain an explicit formula for the distance between …
-
Satellite-Terrestrial Integration under Uncertainty Using Machine Learning and Distributionally Robust Optimization
… ISTNs. To address these issues, we develop a distributionally robust optimization (DRO) model that minimizes the worst-case overall task routing latency under uncertain probability distributions. Utilizing the Wasserstein ambiguity set, the model effectively accommodates unidentified vehicle …
-
Risk-Averse and Distributionally Robust Optimization for Online and Offline Decision-Making
… This dissertation develops risk-sensitive optimization frameworks tailored for environments confronted with uncertainty and distribution shifts in both online and offline settings. </p><p>In the first study, we address the challenge of online stochastic games with risk-averse agents whose …
-
On Multistage Stochastic and Distributionally Robust Optimization: New Algorithms, Complexity Analysis, and Performance Comparison
Multistage optimization under uncertainty refers to sequential decision-making with the presence of uncertainty information that is revealed partially until the end of planning horizon. Depending on the uncertainty model, it is often studied as multistage stochastic optimization (MSO), where one …
-
A Distributionally Robust Optimization Approach for the Optimal Wind Farm Allocation in Multi-Area Power Systems
This dissertation presents a distributionally robust planning model to determine the optimal allocation of wind farms in a multi-area power system, so that the expected energy not served (EENS) is minimized under uncertain conditions of wind power and generator forced outages. Unlike conventional …
-
DATA-DRIVEN ROBUST OPTIMIZATION IN CAPACITY MANAGEMENT
The thesis applies distributionally robust optimization methods in two capacity management projects.
-
Distributionally Robust Chance Constraints for Radiotherapy Treatment Planning under Geometric Uncertainty
… modality for cancer which benefits from optimization. Mainstream approaches define a margin around the tumor called the planning target volume (PTV) to account for changes in patient anatomy throughout the course of treatment. However, the PTV concept does not adequately address …
-
Robust Reinforcement Learning: Theory and Algorithms
… explores novel algorithms in the field of robust reinforcement learning (RL) that address the challenges of controlling dynamical systems in real-world scenarios. Classical reinforcement learning is a powerful sub-field in machine learning for training intelligent sequential decision-making …
-
RISK- AND AMBIGUITY-AVERSE OPTIMIZATION MADE MORE TRACTABLE AND LESS CONSERVATIVE
This thesis builds on the recent developments in distributionally robust optimization and satisficing and proposes how robust decisions under uncertainty can be efficiently made when the decision maker exhibits risk aversion, ambiguity aversion and regret aversion. Theoretically, we develop a new …
-
Robust and Geometry-Aware Machine Learning Using Optimal Transport
… statistical and computational guarantees for robust and geometry-aware learning algorithms. An overarching formalism for this thesis is that of optimal transport (OT), a powerful framework for metric-based comparisons and transformations between high-dimensional data distributions. My work …
-
Federated Learning for Non-Decomposable Objectives
… AUC maximization, partial AUC maximization, distributionally robust optimization, etc. In this thesis, we present our works to address some classes of the federated learning tasks with non-decomposable objectives. In summary, this thesis is four-fold: Firstly, we introduce CODA, which is a …
-
Distributionally robust solution schemes for two-stage optimization and interdiction problems under uncertainty
… in the presence of uncertainty. One can use optimization models with uncertain parameters to formulate the decision problems. Despite its wide applications in real-world problems, optimization under uncertainty gives rise to computational challenges. This thesis aims to design tractable …
-
Distributionally Robust Portfolio Optimization Under Regime-Switching Dynamics
… Systems (rSLDS) within the framework of Distributionally Robust Optimization (DRO) for equity portfolios. In financial markets that are often subject to regime shifts, DRO can be overly conservative when the Wasserstein ambiguity set is required to span multiple regimes. Incorporating a …
-
Stochastic Control and Optimization with Nonstationary Data
… then study out-of-sample performance through a distributionally robust optimization lens, showing that sample-based model predictive control can outperform stochastic dynamic programming when the Bellman operators preserve convexity --- a characteristic of problems with uncertainty appearing in …
-
Market making through reinforcement learning and Wasserstein DRO : an entropy-regularized approach
… including reinforcement learning and Wasserstein distributionally robust optimization (DRO), along with background on classical market making frameworks in financial mathematics. The second part proposes a reinforcement learning framework for options market making, incorporating stochastic …
-
Applications and algorithms for two-stage robust linear optimization
The research scope of this thesis is two-stage robust linear optimization. We are interested in investigating algorithms that can explore its structure and also on adding alternatives to mitigate conservatism inherent to a robust solution. We develop algorithms that incorporate these alternatives …
Page 1 of 4