The Thesis Review [16] Aaron Courville - A Latent Cause Theory of Classical Conditioning
Jan 8, 2021
Aaron Courville, a Professor at the University of Montreal, dives into his PhD thesis on latent cause theory in classical conditioning. He explores the pitfalls of complexity in hypothesis testing, advocating for simplicity. Courville shares his journey from Cornwall to deep learning, discussing how cognitive frameworks shift our understanding of reinforcement. The conversation also touches on generative models and their evolution, alongside the intersection of language and machine learning dynamics, emphasizing the importance of thorough research during one's PhD journey.
AI Snips
Chapters
Transcript
Episode notes
Occam's Razor and Complexity
- Occam's razor favors simpler hypotheses to avoid overfitting and poor generalization.
- However, simpler explanations sometimes mislead when complexity better reflects reality.
Early Career and Research Focus
- Aaron Courville's background was in Engineering Science and biomedical engineering at the University of Toronto.
- His early research involved neural networks and epilepsy treatment models leveraging chaos theory.
Latent Cause Theory of Conditioning
- Classical conditioning is more complex than just reward prediction; animals learn more than reward associations.
- Courville's latent cause theory models animals as inferring latent causes explaining all stimuli, not only reward.