Author name cluster

Daria Terekhov

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

4 papers

2 author rows

NeurIPS Conference 2020 Conference Paper

Learning Linear Programs from Optimal Decisions

Yingcong Tan
Daria Terekhov
Andrew Delong

We propose a flexible gradient-based framework for learning linear programs from optimal decisions. Linear programs are often specified by hand, using prior knowledge of relevant costs and constraints. In some applications, linear programs must instead be learned from observations of optimal decisions. Learning from optimal decisions is a particularly challenging bilevel problem, and much of the related inverse optimization literature is dedicated to special cases. We tackle the general problem, learning all parameters jointly while allowing flexible parameterizations of costs, constraints, and loss functions. We also address challenges specific to learning linear programs, such as empty feasible regions and non-unique optimal decisions. Experiments show that our method successfully learns synthetic linear programs and minimum-cost multi-commodity flow instances for which previous methods are not directly applicable. We also provide a fast batch-mode PyTorch implementation of the homogeneous interior point algorithm, which supports gradients by implicit differentiation or backpropagation.

PDF Details

ICAPS Conference 2013 Conference Paper

Hybrid Queueing Theory and Scheduling Models for Dynamic Environments with Sequence-Dependent Setup Times

Tony T. Tran
Daria Terekhov
Douglas G. Down
J. Christopher Beck

Classically, scheduling research in artificial intelligence has concentrated on the combinatorial challenges arising in a large, static domain where the set of jobs, resource capacities, and other problem parameters are known with certainty and do not change. In contrast, queueing theory has focused primarily on the stochastic arrival and resource requirements of new jobs, de-emphasizing the combinatorics. We study a dynamic parallel scheduling problem with sequence-dependent setup times: arriving jobs must be assigned (online) to one of a set of resources. The jobs have different service times on different resources and there exist setup times that are required to elapse between jobs, depending on both the resource used and the job sequence. We investigate four models that hybridize a scheduling model with techniques from queueing theory to address the dynamic problem. We demonstrate that one of the hybrid models can significantly reduce observed mean flow time performance when compared to the pure scheduling and queueing theory methods. More specifically, at high system loads, our hybrid model achieves a 15% to 60% decrease in mean flow time compared to the pure methodologies. This paper illustrates the advantages of integrating techniques from queueing theory and scheduling to improve performance in dynamic problems with complex combinatorics.

Details

ICAPS Conference 2012 Conference Paper

Long-Run Stability in Dynamic Scheduling

Daria Terekhov
Tony T. Tran
Douglas G. Down
J. Christopher Beck

Stability analysis consists of identifying conditions under which the number of jobs in a system is guaranteed to remain bounded over time. To date, such long-run performance guarantees have not been available for periodic approaches to dynamic scheduling problems. However, stability has been extensively studied in queueing theory. In this paper, we introduce stability to the dynamic scheduling literature and demonstrate that stability guarantees can be obtained for methods that build the schedule for a dynamic problem by periodically solving static deterministic sub-problems. Specifically, we analyze the stability of two dynamic environments: a two-machine flow shop, which has received significant attention in scheduling research, and a polling system with a flow-shop server, an extension of systems typically considered in queueing. We demonstrate that, among stable policies, methods based on periodic optimization of static schedules may achieve better mean flow times than traditional queueing approaches.

Details

AAAI Conference 2007 Conference Paper

Solving a Stochastic Queueing Design and Control Problem with Constraint Programming

Daria Terekhov

A facility with front room and back room operations has the option of hiring specialized or, more expensive, cross-trained workers. Assuming stochastic customer arrival and service times, we seek a smallest-cost combination of cross-trained and specialized workers satisfying constraints on the expected customer waiting time and expected number of workers in the back room. A constraint programming approach using logic-based Benders’ decomposition is presented. Experimental results demonstrate the strong performance of this approach across a wide variety of problem parameters. This paper provides one of the ﬁrst links between queueing optimization problems and constraint programming.

PDF Details