Yingcong Tan Email & Phone Number

Canada

Montreal, Quebec, Canada

Under the supervision of Prof. Chris Beck, my research focused on complex vehicle routing problems (VRP), quadratic unconstrained binary optimization (QUBO) models, and inverse reinforcement learning (IRL). In collaboration with a group of master's and Ph.D. students and post-docs at the TIDEL lab, I worked on the following projects:"Inverse Reinforcement Learning"Study the inverse reinforcement learning problem to learn the discount factor, with an application in animal behaviour study."Complex Vehicle Routing"Study the pickup-and-delivery problem with transfer scheduling (PDPTS), where one must consider the parcel transfer schedule while deciding the delivery routes. "Novel Modeling Scheme for Quadratic Unconstrained Binary Optimization"Study a novel QUBO modelling scheme with applications in Boolean Satisfiability problems, Maximum Independent Set problems and Single-machine Scheduling problems with precedence constraints. "Airplane Transporation for Serving Remote Communities in North Canada"Study an airplane transportation problem which aims to satisfy both passenger and cargo transportation demands in remote communities in northern Canada."Solving Vehicle Routing Problems (VRP) and one-to-one pickup-and-delivery VRP with Quadratic Unconstrained Binary Optimization Models"Manuscripts are either in preparation or under review. Please check https://yingcongtan.github.io/ or https://tidel.mie.utoronto.ca/publications.php for updates on the paper's status.

Montreal, Quebec, Canada

Inverse optimization with Bayesian Optimization.- Learning the objective of integer programming models from (near-)optimal solutions using Bayesian Optimization, the algorithm was applied to solve the Amazon Last-Mile Routing Prediction (e.g., https://routingchallenge.mit.edu/).Data Sampling for Inverse Optimization- Incorporate active learning into inverse optimization to actively select new training data using Bayesian optimization.

Remote

Motivated by the problem of hyper-parameter optimization, the primary focus of my work was to study the necessary convergence conditions of the first-order gradient methods for solving a bi-level optimization problem whose upper-level problem is constrained.

My research focused on learning the coefficients of optimization models from observed optimal solutions, which is also known as the Inverse Optimization (IO) Problem. In particular, we cast IO as a form of machine learning and developed novel gradient algorithms.My work was published at CPAIOr2019 and NeurIPS2020.

Toronto, Ontario, Canada

Toronto, Ontario, Canada

Doctor Of Philosophy - Phd, Industrial Engineering, 4.08/4.3

Activities and Societies: MIAE Graduate Student Committee The Operations Research Challenge Ph.D. thesis on Learning Linear Programs.

Master Of Engineering - Meng, Industrial Engineering, 4.22/4.3

Recipient of the F.A. Gerard Prize, which is awarded annually, when merited, to the most deserving graduate of the Master of/ Magisteriate.

Bachelor Of Applied Science - Basc, Engineering Science

Major in Biomedical Engineering Professional Experience Year: 2010 Sept. - 2011 Aug.