Portrait de Mark Coates

Mark Coates

Membre académique associé
Professeur agrégé, McGill University, Département de génie électrique et informatique
Sujets de recherche
Apprentissage de représentations
Apprentissage sur graphes
Réseaux de neurones en graphes
Systèmes de recommandation
Systèmes dynamiques
Google Scholar

Biographie

Mark Coates est professeur au Département de génie électrique et informatique de l'Université McGill, auquel il s’est joint en 2002. Il a obtenu une licence en génie des systèmes informatiques de l'Université d'Adélaïde (Australie) en 1995 et un doctorat en génie de l'information de l'Université de Cambridge (Royaume-Uni) en 1999. Il a été associé de recherche et conférencier à l'Université Rice, au Texas, de 1999 à 2001. En 2012-2013, il a travaillé en tant que scientifique principal chez Winton Capital Management à Oxford, au Royaume-Uni. Il a assumé de multiples rôles éditoriaux, notamment en tant que rédacteur principal pour IEEE Signal Processing Letters, rédacteur associé pour IEEE Transactions on Signal Processing et rédacteur associé pour IEEE Transactions on Signal and Information Processing over Networks. Les recherches de Mark Coates portent sur l'apprentissage automatique et le traitement statistique des signaux, l'inférence bayésienne et Monte Carlo, et l'apprentissage sur les graphes et les réseaux. Ses contributions les plus influentes et les plus citées concernent la tomographie des réseaux et le filtrage distribué des particules.

Étudiants actuels

Joseph Cotnareanu
Doctorat - McGill
Arcos Bravo David Gamaliel
Maîtrise recherche - McGill
Github
LinkedIn
Google Scholar
Theodore Glavas
Maîtrise recherche - McGill
Google Scholar
Liheng Ma
Doctorat - McGill
Site web
Github
Google Scholar
Julien Nicolas
Doctorat - McGill
regol Regol
Doctorat
Site web
Github
Google Scholar
Chiara Roverato
Doctorat - McGill
Co-superviseur⋅e :
Pablo Piantanida
Antonios Valkanas
Doctorat - McGill
Site web
Google Scholar
Yitian Zhang
Doctorat
Github

Publications

SKOLR: Structured Koopman Operator Linear RNN for Time-Series Forecasting
Yitian Zhang
Liheng Ma
Antonios Valkanas
Boris N. Oreshkin
Mark Coates
Koopman operator theory provides a framework for nonlinear dynamical system analysis and time-series forecasting by mapping dynamics to a sp… (voir plus)ace of real-valued measurement functions, enabling a linear operator representation. Despite the advantage of linearity, the operator is generally infinite-dimensional. Therefore, the objective is to learn measurement functions that yield a tractable finite-dimensional Koopman operator approximation. In this work, we establish a connection between Koopman operator approximation and linear Recurrent Neural Networks (RNNs), which have recently demonstrated remarkable success in sequence modeling. We show that by considering an extended state consisting of lagged observations, we can establish an equivalence between a structured Koopman operator and linear RNN updates. Building on this connection, we present SKOLR, which integrates a learnable spectral decomposition of the input signal with a multilayer perceptron (MLP) as the measurement functions and implements a structured Koopman operator via a highly parallel linear RNN stack. Numerical experiments on various forecasting benchmarks and dynamical systems show that this streamlined, Koopman-theory-based design delivers exceptional performance.
2025-10-05
Proceedings of the 42nd International Conference on Machine Learning (publié)
doi.org
proceedings.mlr.press
Bidirectional Learning for Offline Model-based Biological Sequence Design
Can (Sam) Chen
Yingxue Zhang
Xue Liu
Mark Coates
Offline model-based optimization aims to maximize a black-box objective function with a static dataset of designs and their scores. In this … (voir plus)paper, we focus on biological sequence design to maximize some sequence score. A recent approach employs bidirectional learning, combining a forward mapping for exploitation and a backward mapping for constraint, and it relies on the neural tangent kernel (NTK) of an infinitely wide network to build a proxy model. Though effective, the NTK cannot learn features because of its parametrization, and its use prevents the incorporation of powerful pre-trained Language Models (LMs) that can capture the rich biophysical information in millions of biological sequences. We adopt an alternative proxy model, adding a linear head to a pre-trained LM, and propose a linearization scheme. This yields a closed-form loss and also takes into account the biophysical information in the pre-trained LM. In addition, the forward mapping and the backward mapping play different roles and thus deserve different weights during sequence optimization. To achieve this, we train an auxiliary model and leverage its weak supervision signal via a bi-level optimization framework to effectively learn how to balance the two mappings. Further, by extending the framework, we develop the first learning rate adaptation module \textit{Adaptive}-
2023-07-02
Proceedings of the 40th International Conference on Machine Learning (publié)
doi.org
proceedings.mlr.press