Ján Drgoňa
Pacific Northwest National Laboratory
semi-supervised learning
normalizing flows
optimization
system identification
constrained deep learning
neural state space models
differentiable predictive control
explicit nonlinear mpc
applications
architecture search
dynamical system modeling
3
presentations
SHORT BIO
Jan is a data scientist in the Physics and Computational Sciences Division (PCSD) at Pacific Northwest National Laboratory. His current research focus falls in the intersection of deep learning, constrained optimization, and model-based optimal control. Before joining PNNL, Jan was a postdoc at KU Leuven, Belgium, where he was working on the implementation of model predictive control (MPC) in real-world office buildings. Jan has a PhD in Control Engineering from the Slovak University of Technology in Bratislava, Slovakia. His PhD thesis was on Model Predictive Control with Applications in Building Thermal Comfort Control with the focus on explicit and learning-based MPC.
Presentations
Semi-supervised Learning of Dynamical Systems with Neural Ordinary Differential Equations: A Teacher-Student Model Approach
Yu Wang and 7 other authors
AutoNF: Automated Architecture Optimization of Normalizing Flows with Unconstrained Continuous Relaxation Admitting Optimal Discrete Solution
Yu Wang and 6 other authors
Deep Learning Explicit Differentiable Predictive Control Laws for Buildings
Ján Drgoňa