Scholar
Gido M. van de Ven
Google Scholar ID: 3k0l15MAAAAJ
University of Groningen
Citations & Impact
Citations
4,709
H-index
19
i10-index
22
Publications
20
Co-authors
34
list available
Publications
Resume (English only)
Academic Achievements
- Proposed the influential “three scenarios” framework for continual learning
- Provided proof-of-principle that generative classification is effective for class-incremental learning
- Identified the “stability gap”—a phenomenon where deep neural networks exhibit substantial but temporary forgetting when learning new tasks
- Developed the brain-inspired replay method to mitigate catastrophic forgetting by replaying self-generated abstract memory representations
- Award-winning PhD dissertation
Background
- Assistant Professor at the Bernoulli Institute for Mathematics, Computer Science and Artificial Intelligence, University of Groningen
- Research at the intersection of machine learning, artificial intelligence, and cognitive science
- Aims to understand the computational principles of continual learning—a crucial skill for both artificial and biological agents
- Conducts research through conceptual analysis, computational modeling, deep neural network implementations, and collaborations with experimental labs
- Interested in using insights from neuroscience to make deep neural networks behave more human-like
