TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Neural Ordinary Differential Equations

240 pointsby asparaguiover 6 years ago

5 comments

duvenaudover 6 years ago
Senior author here, I&#x27;m happy to answer any questions.<p>We just released source code: <a href="https:&#x2F;&#x2F;github.com&#x2F;rtqichen&#x2F;torchdiffeq" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;rtqichen&#x2F;torchdiffeq</a> . This includes PyTorch implementations of adaptive ODE solvers that can be differentiated through automatically. So you can mix and match these ODE solvers with any other differentiable model component.<p>There&#x27;s already been a bit of follow-up work, turning Continuous Normalizing Flows into a practical generative density model: <a href="https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;1810.01367" rel="nofollow">https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;1810.01367</a><p>And now we&#x27;re mainly working on 1) Regularizing ODE nets to be faster to solve and 2) getting the time-series model to scale up and extend it to stochastic differential equations.
评论 #18677947 未加载
评论 #18679406 未加载
评论 #18677985 未加载
评论 #18678566 未加载
评论 #18678171 未加载
评论 #18683414 未加载
评论 #18680332 未加载
评论 #18678193 未加载
评论 #18684477 未加载
评论 #18679170 未加载
评论 #18679029 未加载
评论 #18742205 未加载
评论 #18685924 未加载
bitLover 6 years ago
Wow, I need to master ODEs&#x2F;PDEs to keep up with Deep Learning now! Seems like one has to be a master of statistics, operations research, calculus and algorithms to push it forward!<p>Comparison to RNN was impressive! Any well-known real-world models for comparison to state of art?
评论 #18678118 未加载
评论 #18678479 未加载
heinrichfover 6 years ago
Some reporting in MIT&#x27;s newsletter on AI: <a href="https:&#x2F;&#x2F;mailchi.mp&#x2F;technologyreview&#x2F;a-new-type-of-deep-neural-network-that-has-no-layers" rel="nofollow">https:&#x2F;&#x2F;mailchi.mp&#x2F;technologyreview&#x2F;a-new-type-of-deep-neura...</a>
syntaxingover 6 years ago
I remember something similar for CFD application but haven&#x27;t seen much after that. It would be awesome if we can build a cheap and fast Navier Stokes solver with neural networks.
评论 #18678446 未加载
shsjxzhover 6 years ago
Hi, I am very interesting about your models. When you do the back propagation, it seems that it still needs complex calculation. Although O(1) memory cost is an important contribution, do you think record some of the intermediate value will significantly boost the training?