Welcome to my webpage!!

Research Interests

My main research interests concern combinatorial optimization, heuristic search and machine learning, with applications to logistics and supply chain management, production management, resource allocation and information processing.

Explainable AI. I have been conducting extensive research on optimization algorithms for explainable machine learning. We recently proposed a method called Born-again Tree Ensembles which transforms a random forest classifier into the most compact decision tree that has the same decision function.

Optimization for Machine Learning. I am also interested and actively conducting research on combinatorial optimization algorithms (mathematical programming and heuristics) for classical machine learning models, e.g., clustering, community detection, SVM, decision trees, among others. Recently, our HG-means algorithm achieved remarkable clustering results on high-dimensional datasets.

VRP picture
Strategic Problems in Supply Chains Logistics. I study complex optimization problems arising in transportation and logistics, involving strategic decisions with complex daily operations and costs. Several research lines related to this topic are connected to the Unified Hybrid Genetic Search algorithm, which has been tested on over 50 vehicle routing problem variants and represents the current state-of-the-art for this category of problems. I am still actively working on methodological components to improve local-search based heuristics, possibly through Pattern Mining.

Resources and Contact

I strive to make all my academic research results accessible, and publish most source codes on my Github profile. Recent research news, open-source codes and articles are also announced on my Twitter feed, visible on the right side of this page.

On this website, you will find other resources: conference presentations, a list of my published works, along with numerous data sets and results for vehicle routing and other challenging problems. Finally, freel free to contact me. I am more than pleased to discuss recent results, case studies, as well as academic or consulting projects.

Postal adress : Pontifical Catholic University of Rio de Janeiro
Departamento de Informática
Rua Marquês de São Vicente, 225 - Gávea, Rio de Janeiro - RJ, 22451-900, Brazil

Mail to : thibaut.vidal AT cirrelt.ca
or : vidalt AT inf.puc-rio.br