Canada Spends Logo

Explore our Budget 2026 analysis

Grant: Discovery Grants Program - Individual

University of Victoria

$38,000.00

Theory and algorithms for solving bilevel optimization with applications in economics and machine learning

FY 2024-25

Summary
My program of research focuses on studying theories and algorithms for solving the bilevel optimization and its applications from economics, energy market and machine learning. The bilevel  program is an optimization problem where the constraints involve the solution set of another optimization problem.  I plan to  develop applicable optimality conditions and design efficient algorithms with good solution quality for solving bilevel programs. Since bilevel optimization problems are intrinsically non-convex and non-smooth, I will use the variational analysis as a powerful tool. Since a general bilevel program is notoriously difficult to deal with, I will focus on developing theories and algorithm which can solve bilevel programs with some special structures that application areas possessed. I will then apply these general theories to solve  the principal-agent problems in economics, bilevel problems in energy market and the hyper-parameter selection from machine learning.  The principal-agent problem models a fundamental problem that frequently occurs in economics, management science and political science.  It  treats difficulties that arise under conditions of incomplete/asymmetric information when a principal hires an agent to pursue the principal's interests while the agent's action is unobservable. It focuses on designing an incentive scheme with which the principal, seeks to motivate the agent to choose their activities in a way advantageous to the principal. I will try to develop necessary and sufficient optimality condition specially designed for such problem using assumptions which are reasonable in Economics. I will try to find efficient numerical algorithms for solving the principal-agent problem. I will also try to solve  bilevel optimization problems in the energy market. Machine learning focuses on methods and algorithms for computers to learn complex and hidden structures from large data. For this purpose, the learning models construct an optimization problem that can address many learning problems including regression, classification, etc. The resulting optimization typically involves hyper-parameters that must be selected by the user. The selected hyper-parameters determine model accuracy and robustness. Combining the technique we developed for general bilevel programs and the special structures for hyper-parameter selection problems, I hope to develop theory and algorithms for solving these problems. I believe that my research will significantly advance our knowledge of the above proposed problems,  and that the success of my research will help governments, financial institutions, business and industry to make decisions that will produce long-term benefits, such as preventing bankrupcy and even financial crisis. Moreover through working on solving these problems, my HQPs will gain analytic ability to solve  research problems and practical skill and confidence in coping with difficulties that may arise in their future work.
Program
Discovery Grants Program - Individual
Location
British Columbia, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data

As published

Application id
RGPIN-2024-04158
Name nom
Ye, Jane
Department département
Mathematics and Statistics
Organization id
7
Institution établissement
University of Victoria
Province en
British Columbia
Province fr
Colombie-Britannique
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2024
Award amount
38000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1509
Committee name en
Civil, Industrial and Systems Engineering
Committee name fr
Génie civil, industriel et des systèmes
Area of application code
1206
Area of application group en
Advancement of knowledge
Area of application group fr
Avancement des connaissances
Area of application en
Engineering
Area of application fr
Génie
Research subject code
2950
Research subject group en
Applied mathematics
Research subject group fr
Mathématiques appliquées
Research subject en
Applied mathematics
Research subject fr
Mathématiques appliquées
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Theory and algorithms for solving bilevel optimization with applications in economics and machine learning
Keywords
nonsmooth optimization; bilevel optimization; variational analysis; hyperparameter selection; machine learning; principal agent problems; economics; optimality conditions; numerical algorithms; minimax problems
Application summary
My program of research focuses on studying theories and algorithms for solving the bilevel optimization and its applications from economics, energy market and machine learning. The bilevel  program is an optimization problem where the constraints involve the solution set of another optimization problem.  I plan to  develop applicable optimality conditions and design efficient algorithms with good solution quality for solving bilevel programs. Since bilevel optimization problems are intrinsically non-convex and non-smooth, I will use the variational analysis as a powerful tool. Since a general bilevel program is notoriously difficult to deal with, I will focus on developing theories and algorithm which can solve bilevel programs with some special structures that application areas possessed. I will then apply these general theories to solve  the principal-agent problems in economics, bilevel problems in energy market and the hyper-parameter selection from machine learning.  The principal-agent problem models a fundamental problem that frequently occurs in economics, management science and political science.  It  treats difficulties that arise under conditions of incomplete/asymmetric information when a principal hires an agent to pursue the principal's interests while the agent's action is unobservable. It focuses on designing an incentive scheme with which the principal, seeks to motivate the agent to choose their activities in a way advantageous to the principal. I will try to develop necessary and sufficient optimality condition specially designed for such problem using assumptions which are reasonable in Economics. I will try to find efficient numerical algorithms for solving the principal-agent problem. I will also try to solve  bilevel optimization problems in the energy market. Machine learning focuses on methods and algorithms for computers to learn complex and hidden structures from large data. For this purpose, the learning models construct an optimization problem that can address many learning problems including regression, classification, etc. The resulting optimization typically involves hyper-parameters that must be selected by the user. The selected hyper-parameters determine model accuracy and robustness. Combining the technique we developed for general bilevel programs and the special structures for hyper-parameter selection problems, I hope to develop theory and algorithms for solving these problems. I believe that my research will significantly advance our knowledge of the above proposed problems,  and that the success of my research will help governments, financial institutions, business and industry to make decisions that will produce long-term benefits, such as preventing bankrupcy and even financial crisis. Moreover through working on solving these problems, my HQPs will gain analytic ability to solve  research problems and practical skill and confidence in coping with difficulties that may arise in their future work.
University of Victoria - $38,000 Grant (2024-25) | Canada Spends