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.

Neuro Evolution of Augmented Topologies

95 pointsby charlieirishover 3 years ago

5 comments

pkoperekover 3 years ago
Apart from NEAT&#x2F;HyperNEAT there are also other approaches to neuroevolution (I think in this context it is referred to as &quot;Evolutionary Neural Architecture Search&quot; [0]). Evolution in general can be applied in different ways (e.g. optimizing the architecture, replacing training gradient descent etc.).<p>A while ago I co-authored a paper in this space [1] and released some code for interested folks [2].<p>[0]: <a href="https:&#x2F;&#x2F;arxiv.org&#x2F;pdf&#x2F;2008.10937.pdf" rel="nofollow">https:&#x2F;&#x2F;arxiv.org&#x2F;pdf&#x2F;2008.10937.pdf</a><p>[1]: <a href="https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;1801.00119" rel="nofollow">https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;1801.00119</a><p>[2]: <a href="https:&#x2F;&#x2F;gitlab.com&#x2F;pkoperek&#x2F;pytorch-dnn-evolution&#x2F;-&#x2F;tree&#x2F;master" rel="nofollow">https:&#x2F;&#x2F;gitlab.com&#x2F;pkoperek&#x2F;pytorch-dnn-evolution&#x2F;-&#x2F;tree&#x2F;mas...</a>
评论 #29332794 未加载
eggyover 3 years ago
They seem to be similar to Gene Sher&#x27;s TWEANNS - Topology and Weight Evolving Neural Networks - that I learned about in his 2012 book, &quot;Handbook of Neuroevolution Through Erlang&quot; (sure, it didn&#x27;t pick up because Erlang ;))<p>Gene&#x27;s specific implementation is DXNN (Discover and eXplore Neural Network) implemented in 2010 [1]<p>[1] <a href="https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;1008.2412" rel="nofollow">https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;1008.2412</a>
评论 #29330554 未加载
charles_fover 3 years ago
Author there, I&#x27;m curious how you found this post that was not indexed yet? :) (and not finished then)
MrQuincleover 3 years ago
There&#x27;s also for example HyerNEAT. I&#x27;m out of the field now. Has recent progress been made recently with these technologies?
评论 #29328299 未加载
optimalsolverover 3 years ago
For the logical endpoint of this approach, check out FreeWire:<p><a href="https:&#x2F;&#x2F;github.com&#x2F;noahtren&#x2F;Freewire" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;noahtren&#x2F;Freewire</a><p>You can experiment with freely wired neural networks without traditional layers.