Daria Terekhov
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Daria Terekhov has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Taewoo Lee
University of Pittsburgh
7 shared publications - Timothy C. Y. Chan
Department of Mechanical and Industrial Engineering, University of Toronto, Toronto, Ontario M5S 3G8, Canada
4 shared publications - Aaron Babier
Department of Mechanical and Industrial Engineering, University of Toronto, Toronto, Ontario M5S 3G8, Canada
2 shared publications - Rafid Mahmood
Telfer School of Management, University of Ottawa, Ottawa, Ontario K1N 6N5, Canada
2 shared publications - Andrew Delong
University of Western Ontario
1 shared publication - Daniel M. Frances
Department of Mechanical and Industrial Engineering, University of Toronto, Toronto, Ontario, Canada, M5S 3G8
1 shared publication
Latest Publications
- 30 September 2022
- Published by Springer Nature
- p. 1-5
Inverse Integer Optimization
- 30 September 2022
- Published by Springer Nature
- p. 1-6
Inverse Optimization
- 30 September 2022
- Published by Springer Nature
- p. 1-5
Data-Driven Inverse Optimization
- 25 January 2022
- Published by Wiley in International Transactions in Operational Research
- Vol. 29 (6) , 3514-3536
Inverse attribute‐based optimization with an application in assortment optimization
Many applications of inverse optimization (IO) arise in settings where the goal is to predict the future actions of an optimizing agent (e.g., an optimizing customer's future purchases). The majority of papers in this area implicitly assume an alternative‐based modeling approach: The forward model finds an optimal set of actions (decisions) from among a given set of alternatives, while the inverse model imputes objective function coefficients corresponding to these alternatives. Since the imputed weights correspond only to alternatives existing in the training set, alternative‐based modeling is limited to applications where the set of options does not change when a prediction is needed. In this paper, we apply an attribute‐based perspective, which allows IO to impute the weights of attributes that lead to an optimal decision instead of imputing the weight of the decision itself. This perspective expands the range of IO applicability; we demonstrate that it facilitates the application of IO in assortment optimization, where changing product selections is a defining feature and accurate predictions of demand are important. We compare inverse attribute‐based optimization with rank‐based and machine learning methods. We show that since IO encodes the utility optimizing behavior of the consumer into the preference learning process, it results in lower assortment regret for the store and a lower utility gap for the consumers.
- 25 January 2022
- Published by Wiley in International Transactions in Operational Research
- Vol. 29 (6) , 3514-3536
Inverse attribute‐based optimization with an application in assortment optimization
Many applications of inverse optimization (IO) arise in settings where the goal is to predict the future actions of an optimizing agent (e.g., an optimizing customer's future purchases). The majority of papers in this area implicitly assume an alternative‐based modeling approach: The forward model finds an optimal set of actions (decisions) from among a given set of alternatives, while the inverse model imputes objective function coefficients corresponding to these alternatives. Since the imputed weights correspond only to alternatives existing in the training set, alternative‐based modeling is limited to applications where the set of options does not change when a prediction is needed. In this paper, we apply an attribute‐based perspective, which allows IO to impute the weights of attributes that lead to an optimal decision instead of imputing the weight of the decision itself. This perspective expands the range of IO applicability; we demonstrate that it facilitates the application of IO in assortment optimization, where changing product selections is a defining feature and accurate predictions of demand are important. We compare inverse attribute‐based optimization with rank‐based and machine learning methods. We show that since IO encodes the utility optimizing behavior of the consumer into the preference learning process, it results in lower assortment regret for the store and a lower utility gap for the consumers.
- 23 March 2021
- Published by Springer Nature in Optimization Letters
- Vol. 15 (4) , 1061-1079
Inverse integer optimization with multiple observations
- 23 March 2021
- Published by Springer Nature in Optimization Letters
- Vol. 15 (4) , 1061-1079
Inverse integer optimization with multiple observations
- 1 January 2021
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in INFORMS Journal on Optimization
- Vol. 3 (2) , 119-138
An Ensemble Learning Framework for Model Fitting and Evaluation in Inverse Linear Optimization
We develop a generalized inverse optimization framework for fitting the cost vector of a single linear optimization problem given multiple observed decisions. This setting is motivated by ensemble learning, where building consensus from base learners can yield better predictions. We unify several models in the inverse optimization literature under a single framework and derive assumption-free and exact solution methods for each one. We extend a goodness-of-fit metric previously introduced for the problem with a single observed decision to this new setting and demonstrate several important properties. Finally, we demonstrate our framework in a novel inverse optimization-driven procedure for automated radiation therapy treatment planning. Here, the inverse optimization model leverages an ensemble of dose predictions from different machine learning models to construct a consensus treatment plan that outperforms baseline methods. The consensus plan yields better trade-offs between the competing clinical criteria used for plan evaluation.
- 1 January 2021
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in INFORMS Journal on Optimization
- Vol. 3 (2) , 119-138
An Ensemble Learning Framework for Model Fitting and Evaluation in Inverse Linear Optimization
We develop a generalized inverse optimization framework for fitting the cost vector of a single linear optimization problem given multiple observed decisions. This setting is motivated by ensemble learning, where building consensus from base learners can yield better predictions. We unify several models in the inverse optimization literature under a single framework and derive assumption-free and exact solution methods for each one. We extend a goodness-of-fit metric previously introduced for the problem with a single observed decision to this new setting and demonstrate several important properties. Finally, we demonstrate our framework in a novel inverse optimization-driven procedure for automated radiation therapy treatment planning. Here, the inverse optimization model leverages an ensemble of dose predictions from different machine learning models to construct a consensus treatment plan that outperforms baseline methods. The consensus plan yields better trade-offs between the competing clinical criteria used for plan evaluation.
- 20 November 2020
- Published by Springer Nature in Health Care Management Science
- Vol. 24 (1) , 26-40
A robust optimization model for tactical capacity planning in an outpatient setting
Tactical capacity planning is a key element of planning and control decisions in healthcare settings, focusing on the medium-term allocation of a clinic’s resources to appointments of different types. One of the most scarce resources in healthcare is physician time. Due to uncertainty in demand for appointments, it is difficult to provide an exact match between the planned physician availability and appointment requests. Our study uses cardinality-constrained robust optimization to develop tactical capacity plans which are robust against uncertainty, providing a feasible allocation of capacity for all realizations of demand to the extent allowed by the budget of uncertainty. The outpatient setting we consider sees first-visit patients and re-visit patients, and both patient types have access time targets. We experimentally evaluate our robust model and its practical implications under different levels of conservatism. We show that we can guarantee 100% feasibility of the robust tactical capacity plan while not being fully conservative, which will lead to the clinic saving money while being able to meet demand despite uncertainty. We also show how the robust model helps us to identify the critical time periods leading to worst case physician peak load, which could be valuable to decision-makers. Throughout the experiments, we find that the step of translating available data into an uncertainty set can influence the true conservatism of a solution.
- 20 November 2020
- Published by Springer Nature in Health Care Management Science
- Vol. 24 (1) , 26-40
A robust optimization model for tactical capacity planning in an outpatient setting
Tactical capacity planning is a key element of planning and control decisions in healthcare settings, focusing on the medium-term allocation of a clinic’s resources to appointments of different types. One of the most scarce resources in healthcare is physician time. Due to uncertainty in demand for appointments, it is difficult to provide an exact match between the planned physician availability and appointment requests. Our study uses cardinality-constrained robust optimization to develop tactical capacity plans which are robust against uncertainty, providing a feasible allocation of capacity for all realizations of demand to the extent allowed by the budget of uncertainty. The outpatient setting we consider sees first-visit patients and re-visit patients, and both patient types have access time targets. We experimentally evaluate our robust model and its practical implications under different levels of conservatism. We show that we can guarantee 100% feasibility of the robust tactical capacity plan while not being fully conservative, which will lead to the clinic saving money while being able to meet demand despite uncertainty. We also show how the robust model helps us to identify the critical time periods leading to worst case physician peak load, which could be valuable to decision-makers. Throughout the experiments, we find that the step of translating available data into an uncertainty set can influence the true conservatism of a solution.
- Open Access
An Ensemble Learning Framework for Model Fitting and Evaluation in Inverse Linear Optimization
We develop a generalized inverse optimization framework for fitting the cost vector of a single linear optimization problem given multiple observed decisions. This setting is motivated by ensemble learning, where building consensus from base learners can yield better predictions. We unify several models in the inverse optimization literature under a single framework and derive assumption-free and exact solution methods for each one. We extend a goodness-of-fit metric previously introduced for the problem with a single observed decision to this new setting, and demonstrate several important properties. Finally, we demonstrate our framework in a novel inverse optimization-driven procedure for automated radiation therapy treatment planning. Here, the inverse optimization model leverages an ensemble of dose predictions from different machine learning models to construct a consensus treatment plan that outperforms baseline methods. The consensus plan yields better trade-offs between the competing clinical criteria used for plan evaluation.
Open AccessAn Ensemble Learning Framework for Model Fitting and Evaluation in Inverse Linear Optimization
We develop a generalized inverse optimization framework for fitting the cost vector of a single linear optimization problem given multiple observed decisions. This setting is motivated by ensemble learning, where building consensus from base learners can yield better predictions. We unify several models in the inverse optimization literature under a single framework and derive assumption-free and exact solution methods for each one. We extend a goodness-of-fit metric previously introduced for the problem with a single observed decision to this new setting, and demonstrate several important properties. Finally, we demonstrate our framework in a novel inverse optimization-driven procedure for automated radiation therapy treatment planning. Here, the inverse optimization model leverages an ensemble of dose predictions from different machine learning models to construct a consensus treatment plan that outperforms baseline methods. The consensus plan yields better trade-offs between the competing clinical criteria used for plan evaluation.
- 1 March 2019
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 65 (3) , 1115-1135
Inverse Optimization: Closed-Form Solutions, Geometry, and Goodness of Fit
In classical inverse linear optimization, one assumes that a given solution is a candidate to be optimal. Real data are imperfect and noisy, so there is no guarantee that this assumption is satisfied. Inspired by regression, this paper presents a unified framework for cost function estimation in linear optimization comprising a general inverse optimization model and a corresponding goodness-of-fit metric. Although our inverse optimization model is nonconvex, we derive a closed-form solution and present the geometric intuition. Our goodness-of-fit metric, ρ, the coefficient of complementarity, has similar properties to R2 from regression and is quasi-convex in the input data, leading to an intuitive geometric interpretation. While ρ is computable in polynomial time, we derive a lower bound that possesses the same properties, is tight for several important model variations, and is even easier to compute. We demonstrate the application of our framework for model estimation and evaluation in production planning and cancer therapy. This paper was accepted by Yinyu Ye, optimization.
- 1 March 2019
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 65 (3) , 1115-1135
Inverse Optimization: Closed-Form Solutions, Geometry, and Goodness of Fit
In classical inverse linear optimization, one assumes that a given solution is a candidate to be optimal. Real data are imperfect and noisy, so there is no guarantee that this assumption is satisfied. Inspired by regression, this paper presents a unified framework for cost function estimation in linear optimization comprising a general inverse optimization model and a corresponding goodness-of-fit metric. Although our inverse optimization model is nonconvex, we derive a closed-form solution and present the geometric intuition. Our goodness-of-fit metric, ρ, the coefficient of complementarity, has similar properties to R2 from regression and is quasi-convex in the input data, leading to an intuitive geometric interpretation. While ρ is computable in polynomial time, we derive a lower bound that possesses the same properties, is tight for several important model variations, and is even easier to compute. We demonstrate the application of our framework for model estimation and evaluation in production planning and cancer therapy. This paper was accepted by Yinyu Ye, optimization.
- Open Access
- 1 January 2019
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in INFORMS Transactions on Education
- Vol. 19 (2) , 67-80
A Case-Based Undergraduate Operations Research Course
This paper describes a purely case-based undergraduate operations research course that has been successfully run in the industrial engineering program at the University of Toronto for 11 years. We describe the structure of the course, including the process adopted for solving the weekly cases, student assessment methods, and the choice of cases. Student feedback suggests that the course enhances the learning experience through the principle of learning by doing, provides a platform for integration of operations research methods learned in earlier classes, can be conducive to improving communication skills, and addresses the gap between theory and practice through a simulated workplace environment.
- 1 January 2019
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in INFORMS Transactions on Education
- Vol. 19 (2) , 67-80
Open AccessA Case-Based Undergraduate Operations Research Course
This paper describes a purely case-based undergraduate operations research course that has been successfully run in the industrial engineering program at the University of Toronto for 11 years. We describe the structure of the course, including the process adopted for solving the weekly cases, student assessment methods, and the choice of cases. Student feedback suggests that the course enhances the learning experience through the principle of learning by doing, provides a platform for integration of operations research methods learned in earlier classes, can be conducive to improving communication skills, and addresses the gap between theory and practice through a simulated workplace environment.