Discover and read the best of Twitter Threads about #generalization
Most recents (2)
Yann LeCun, a true pioneer of AI, emphasizes the importance of Open Science in driving progress in AI. Let's dive into his insights on the Future of AI and why Open Science is crucial for its development. #YannLeCun #OpenScience #FutureOfAI 
🧐LeCun believes that the future of AI lies in developing systems that can learn from fewer examples and generalize better. This requires a deeper understanding of the underlying principles of intelligence. #AI #Generalization #Intelligence
🤝LeCun created FAIR @MetaAI, a research organization within Meta, to make long-term investments in AI research at scale. He emphasizes the importance of open science in AI research, where researchers share their data and findings openly to accelerate progress. #Fair #OpenScience
Can we infuse #structure into a time series (#TS) model from a diverse dataset so as to greatly improve #generalization on new TS coming from different datasets?
Yes, via a new principle called #Representational Time-Frequency Consistency (TF-C)
arxiv.org/abs/2206.08496 1/3
Yes, via a new principle called #Representational Time-Frequency Consistency (TF-C)
arxiv.org/abs/2206.08496 1/3
@xiangzhang1015 @MITLL @Harvard @HarvardDBMI Results suggest models can be trained on a source TS dataset and deployed on a range of target TS datasets without retraining, resulting in performance that is better than that of state-of-the-art & state-of-practice models 2/3
@xiangzhang1015 @MITLL @Harvard @HarvardDBMI Paper: arxiv.org/abs/2206.08496
Data: zitniklab.hms.harvard.edu/projects/TF-C
Code: github.com/mims-harvard/T…
Enjoyed working on this project, which is led by postdoc @xiangzhang1015 @HarvardDBMI with invaluable help from @Harvard undergraduates and @MITLL collaborators 3/3
Data: zitniklab.hms.harvard.edu/projects/TF-C
Code: github.com/mims-harvard/T…
Enjoyed working on this project, which is led by postdoc @xiangzhang1015 @HarvardDBMI with invaluable help from @Harvard undergraduates and @MITLL collaborators 3/3
