Jump to content

Talk:Hierarchical Deep Learning

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Why was this article marked for deletion?

[edit]

I don't understand -- even if the current state of the article shows that it may be written by more than 1 author, there is a substantial body of research on the topic that is also well-cited, at least back to 2012:

- Luo, Ping, Xiaogang Wang, and Xiaoou Tang. "Hierarchical face parsing via deep learning." Computer Vision and Pattern Recognition (CVPR), 2012 IEEE Conference on. IEEE, 2012.

- Kulkarni, Tejas D., et al. "Hierarchical deep reinforcement learning: Integrating temporal abstraction and intrinsic motivation." Advances in neural information processing systems. 2016.

- Huang, Gary B., Honglak Lee, and Erik Learned-Miller. "Learning hierarchical representations for face verification with convolutional deep belief networks." Computer Vision and Pattern Recognition (CVPR), 2012 IEEE Conference on. IEEE, 2012.

If anything, this should be a candidate for revision, but not deletion. -Atasato (talk) 20:43, 9 June 2018 (UTC)[reply]

Edit: Upon closer inspection, there is already reinforcement learning, of which deep reinforcement learning is a subtopic of. I suppose hierarchical deep learning is just a sub-technique under this. -Atasato (talk) 20:49, 9 June 2018 (UTC)[reply]

@Atasato: as you can easily see, this article is actually about "HDLTex", and written by the author of HDLTex... the sources you mention do not apply to this articles contents. At the minimum, it needs to be cleaned up for this obvious WP:SELFCITE and WP:COI. Or deleted, as "hierarchical deep learning" is probably not a proper well-defined topic, but there are just many very different hierarchical deep things. HelpUsStopSpam (talk) 22:13, 9 June 2018 (UTC)[reply]
@HelpUsStopSpam: Perhaps it would be best to simply move the contents of this page that we can abstract to the main page on deep learning or reinforcement learning? I think it's worth at least a small section or a bullet -- would it be okay to just leave it as a short stub and flag it as an "ongoing research" article or something? (I'm new to this editing thing so please excuse me.) -Atasato (talk) 22:19, 9 June 2018 (UTC)[reply]
@Atasato: as said before, I am not sure it is a well-defined topic. Isn't every deep convolutional model somehow "hierarchical"? And the current article doesn't have contents worth keeping for integration in the existing articles - either they are failing independent review, or they are already in the other articles... HelpUsStopSpam (talk) 11:24, 10 June 2018 (UTC)[reply]
Since the deep learning article begins with "Deep learning (also known as deep structured learning or hierarchical learning) is ..." I made this simply a redirect to deep learning = hierarchical learning. HelpUsStopSpam (talk) 11:41, 10 June 2018 (UTC)[reply]
I agree with HelpUsStopSpam's reasoning that this article is not clearly about a topic that we do not already cover, and so by WP:IAR I'm not going to query this at WP:DRV, but this isn't how things are supposed to work: firstly, I don't consider this to be clearly spam, rather the problem with it is that by our standards it is a WP:NEOLOGISM and so the content cannot avoid being Wikipedia:Original research, so the rationale isn't quite right; secondly, in removing the PROD notice and asking for explanation here, Atasato did query the PROD, therefore it should have gone to AfD (PRODs really need self-explanatory rationales; if they require further justification, the PROD should fail); third, PRODs should be closed by admins, who are supposedly the users who have the working support of the community to execute deletion policy; finally, not a hard point of policy, but it is courteous to inform MatthewVanitas, who had reviewed and passed the article at WP:AfC, and, assuming good faith, the article creator, Kamykowsari. — Charles Stewart (talk) 08:06, 12 April 2021 (UTC)[reply]