Search for tag: "generalization"

3 Media

How Nature’s Tiniest Soda Could Help Feed… duration 3 minutes 5 seconds How Nature’s Tiniest Soda Could Help Feed… How Nature’s Tiniest Soda Could Help Feed the World: The Story of the Algal Pyrenoid, Eric Franklin, G4 (3965535)	How Nature’s Tiniest Soda Could Help Feed the World: The Story of the Algal Pyrenoid, Eric Franklin, G4 (3965535) Plants are responsible for almost everything around us. Not just as crops for us or livestock to eat, but also for the wood that makes up our houses and furniture, the cotton that we wear as clothes,… reinforcement learning generalization interpretability neural network tubules co2 algae 3d plants proteins soda sugar membranes crop nature reconstructions difference rest middle matrix fertilizer water yields food From Research Princeton Research Day 6 likes 288 plays Date April 27th, 2022 Speaker Eric Franklin Department Molecular Biology
Using a Theory of Human Cognition to Improve… duration 3 minutes 0 seconds Using a Theory of Human Cognition to Improve… Using a Theory of Human Cognition to Improve Computer Agents, Albert Lin, UG '23 (3965329)	Using a Theory of Human Cognition to Improve Computer Agents, Albert Lin, UG '23 (3965329) Current reinforcement learning (RL) models still have difficulty generalizing to novel, but related, tasks. They are also often composed of deep neural networks which carry out computations that do… reinforcement learning generalization interpretability neural network model computer humans game baseline theory control attention levels models visualizations experiments network approach training decisions cognition initializations fruits From Research Princeton Research Day 3 likes 66 plays Date April 27th, 2022 Speaker Albert Lin Department Computer Science
Bartlett, Peter "Optimization and… duration 44 minutes 47 seconds Bartlett, Peter "Optimization and… Bartlett, Peter "Optimization and Generalization Properties of Deep Neural Networks" May 14, 2018	Bartlett, Peter "Optimization and Generalization Properties of Deep Neural Networks" May 14, 2018 PCTS "Bridging Mathematical Optimization Information Theory and Data Science" pcts peter bartlett optimization and generalization properties of deep neural networks captions_12_17_2020 identity linear function functions things networks composition gradient matrix result network step matrices depth vector criterion thing condition sort instance From Broadcast Broadcast Center Staff 0 likes 167 plays Date May 14th, 2018 Speaker Peter Bartlett Department PCTS Location 407 Jadwin