Talk:1838: Machine Learning - Revision history

172.70.47.58: LLM-Era References

2024-02-21T03:19:46Z

LLM-Era References

172.70.117.76 at 22:56, 8 June 2021

2021-06-08T22:56:28Z

Singlelinelabyrinth at 05:22, 27 July 2020

2020-07-27T05:22:36Z

Misterstick: /* Fixing the explanation */ seconded

2018-04-05T11:29:20Z

‎Fixing the explanation: seconded

172.68.47.18 at 21:08, 24 August 2017

2017-08-24T21:08:24Z

Hkmaly at 03:20, 27 May 2017

2017-05-27T03:20:50Z

162.158.106.42: Names

2017-05-24T08:38:50Z

Names

162.158.122.66: Never edit a comment you didn't make. That's not how discussions work.

2017-05-22T19:07:05Z

Never edit a comment you didn't make. That's not how discussions work.

172.68.54.112 at 17:46, 20 May 2017

2017-05-20T17:46:20Z

172.68.54.112 at 17:45, 20 May 2017

2017-05-20T17:45:37Z

← Older revision		Revision as of 03:19, 21 February 2024
Line 43:		Line 43:

	I like the use of "Cueball Prime" and "Cueball II" in the transcript. can someone make that a part of the style guide?		I like the use of "Cueball Prime" and "Cueball II" in the transcript. can someone make that a part of the style guide?
		+
		+	====LLM-Era References (2023+)====
		+	This was just reposted on the "Marcus On AI" Substack, forwarded through Dave Farber's "Interesting People" list, in reference to LLM ChatGPT allegedly "going beserk".
		+	https://garymarcus.substack.com/p/chatgpt-has-gone-berserk
		+	https://ip.topicbox.com/groups/ip/T775bf6cac82b32a1/chatgpt-has-gone-berserk
		+	[[Special:Contributions/172.70.47.58\|172.70.47.58]] 03:19, 21 February 2024 (UTC)

← Older revision		Revision as of 22:56, 8 June 2021
Line 41:		Line 41:
	:seconded --[[User:Misterstick\|Misterstick]] ([[User talk:Misterstick\|talk]]) 11:29, 5 April 2018 (UTC)		:seconded --[[User:Misterstick\|Misterstick]] ([[User talk:Misterstick\|talk]]) 11:29, 5 April 2018 (UTC)
	::I'm a machine learning geek and I'll try to add an explanation of "recurrent", a few minutes. [[User:Singlelinelabyrinth\|Singlelinelabyrinth]] ([[User talk:Singlelinelabyrinth\|talk]]) 05:22, 27 July 2020 (UTC)		::I'm a machine learning geek and I'll try to add an explanation of "recurrent", a few minutes. [[User:Singlelinelabyrinth\|Singlelinelabyrinth]] ([[User talk:Singlelinelabyrinth\|talk]]) 05:22, 27 July 2020 (UTC)
		+
		+	I like the use of "Cueball Prime" and "Cueball II" in the transcript. can someone make that a part of the style guide?

← Older revision		Revision as of 05:22, 27 July 2020
Line 40:		Line 40:
	I only came here to get an explanation of "recurrent," and I canʻt find it.		I only came here to get an explanation of "recurrent," and I canʻt find it.
	:seconded --[[User:Misterstick\|Misterstick]] ([[User talk:Misterstick\|talk]]) 11:29, 5 April 2018 (UTC)		:seconded --[[User:Misterstick\|Misterstick]] ([[User talk:Misterstick\|talk]]) 11:29, 5 April 2018 (UTC)
		+	::I'm a machine learning geek and I'll try to add an explanation of "recurrent", a few minutes. [[User:Singlelinelabyrinth\|Singlelinelabyrinth]] ([[User talk:Singlelinelabyrinth\|talk]]) 05:22, 27 July 2020 (UTC)

← Older revision		Revision as of 11:29, 5 April 2018
Line 39:		Line 39:

	I only came here to get an explanation of "recurrent," and I canʻt find it.		I only came here to get an explanation of "recurrent," and I canʻt find it.
		+	:seconded --[[User:Misterstick\|Misterstick]] ([[User talk:Misterstick\|talk]]) 11:29, 5 April 2018 (UTC)

← Older revision		Revision as of 21:08, 24 August 2017
Line 37:		Line 37:
	====Fixing the explanation====		====Fixing the explanation====
	Right now, the explanation has two parts, one that is simply trying to explain it for the casual reader, and another that goes into the details of machine/deep learning, linear algebra, neural networks etc. (I almost forgot composting!) The way the two parts are jumbled together makes no sense. Perhaps having a simple initial explanation with subsections for more detailed explanation of individual topics relevant to the comic would fix the mess. [[User:Nialpxe\|Nialpxe]] ([[User talk:Nialpxe\|talk]]) 14:08, 19 May 2017 (UTC)		Right now, the explanation has two parts, one that is simply trying to explain it for the casual reader, and another that goes into the details of machine/deep learning, linear algebra, neural networks etc. (I almost forgot composting!) The way the two parts are jumbled together makes no sense. Perhaps having a simple initial explanation with subsections for more detailed explanation of individual topics relevant to the comic would fix the mess. [[User:Nialpxe\|Nialpxe]] ([[User talk:Nialpxe\|talk]]) 14:08, 19 May 2017 (UTC)
		+
		+	I only came here to get an explanation of "recurrent," and I canʻt find it.

← Older revision		Revision as of 03:20, 27 May 2017
Line 18:		Line 18:
	The comment that SVMs would be a better paradigm, rather than neural networks, is kind of wrong. Anyone who's worked with neural networks knows they're still essentially a linear algebra problem, just with nonlinear activation functions. Play around with tensorflow (it's fun and educational!) and you'll find most of the linear algebra isn't abstracted away as it might be in Keras, SkLearn or Caret (R). That being said, interpretability is absolutely a problem with these complex models. This is as much because the world doesn't like conforming to the nice modernist notion of a sensible theory (ie. one that can be reduced to a nice linear relationship), but even things like L1 regularisation often leave you wondering "but how does it all fit together?". On the other hand, while methods like SVMs still have a bit of machine learning magic in resolving how its hyperplane divides the hyperspace (ie. the value is derived empirically, not theoretically), the results are typically human interpretable, for a given definition of interpretable. It's no y= wx + b, but it's definitely possible. Same same for most methods short of very deep neural nets with millions of parameters. Most machine learning experts I've met have a pretty good idea what is going on in the simpler models, such as CARTs, SVMs, boosted models etc. The only reason neural nets are blackbox-y is that there's a huge amount going on inside them, and it's too much effort to do more than analyse outputs! [[Special:Contributions/172.68.141.142\|172.68.141.142]] 22:43, 17 May 2017 (UTC)		The comment that SVMs would be a better paradigm, rather than neural networks, is kind of wrong. Anyone who's worked with neural networks knows they're still essentially a linear algebra problem, just with nonlinear activation functions. Play around with tensorflow (it's fun and educational!) and you'll find most of the linear algebra isn't abstracted away as it might be in Keras, SkLearn or Caret (R). That being said, interpretability is absolutely a problem with these complex models. This is as much because the world doesn't like conforming to the nice modernist notion of a sensible theory (ie. one that can be reduced to a nice linear relationship), but even things like L1 regularisation often leave you wondering "but how does it all fit together?". On the other hand, while methods like SVMs still have a bit of machine learning magic in resolving how its hyperplane divides the hyperspace (ie. the value is derived empirically, not theoretically), the results are typically human interpretable, for a given definition of interpretable. It's no y= wx + b, but it's definitely possible. Same same for most methods short of very deep neural nets with millions of parameters. Most machine learning experts I've met have a pretty good idea what is going on in the simpler models, such as CARTs, SVMs, boosted models etc. The only reason neural nets are blackbox-y is that there's a huge amount going on inside them, and it's too much effort to do more than analyse outputs! [[Special:Contributions/172.68.141.142\|172.68.141.142]] 22:43, 17 May 2017 (UTC)

		+	:I remember from school that neural nets can get extremely hard to analyze even when they only contain five neurons. -- [[User:Hkmaly\|Hkmaly]] ([[User talk:Hkmaly\|talk]]) 03:20, 27 May 2017 (UTC)

	Does anyone else think the topic may have been influenced by Google's recently (May 17) featured article about machine learning?[[https://www.google.com/intl/en/about/main/gender-equality-films/]]		Does anyone else think the topic may have been influenced by Google's recently (May 17) featured article about machine learning?[[https://www.google.com/intl/en/about/main/gender-equality-films/]]

← Older revision		Revision as of 08:38, 24 May 2017
Line 30:		Line 30:

	The fuck is "Pinball"? [[Special:Contributions/162.158.122.66\|162.158.122.66]] 03:59, 19 May 2017 (UTC)		The fuck is "Pinball"? [[Special:Contributions/162.158.122.66\|162.158.122.66]] 03:59, 19 May 2017 (UTC)
		+	: Agree, expunged it. [[Special:Contributions/162.158.106.42\|162.158.106.42]] 08:38, 24 May 2017 (UTC)

	On the topic of 'Stirring', I'm not sure why it's being associated with neural networks. It's a common thing in machine learning to randomize starting conditions to avoid local minima. This does exist in neural networks, as edge weights are typically randomized, but it's also the first step in many different algorithms, such as k-means where the initial centroid locations are randomized, or decision trees where random forests are sometimes used. [[Special:Contributions/173.245.50.186\|173.245.50.186]] 13:18, 19 May 2017 (UTC) sbendl		On the topic of 'Stirring', I'm not sure why it's being associated with neural networks. It's a common thing in machine learning to randomize starting conditions to avoid local minima. This does exist in neural networks, as edge weights are typically randomized, but it's also the first step in many different algorithms, such as k-means where the initial centroid locations are randomized, or decision trees where random forests are sometimes used. [[Special:Contributions/173.245.50.186\|173.245.50.186]] 13:18, 19 May 2017 (UTC) sbendl

@@ Line 29: / Line 29: @@
 :: I lovingly think of this site as "Over-Explain XKCD"[[Special:Contributions/172.68.54.112|172.68.54.112]] 17:44, 20 May 2017 (UTC)
-The fuck is "{{w|Pinball}}"? [[Special:Contributions/162.158.122.66|162.158.122.66]] 03:59, 19 May 2017 (UTC)
+The fuck is "Pinball"? [[Special:Contributions/162.158.122.66|162.158.122.66]] 03:59, 19 May 2017 (UTC)
 On the topic of 'Stirring', I'm not sure why it's being associated with neural networks. It's a common thing in machine learning to randomize starting conditions to avoid local minima. This does exist in neural networks, as edge weights are typically randomized, but it's also the first step in many different algorithms, such as k-means where the initial centroid locations are randomized, or decision trees where random forests are sometimes used. [[Special:Contributions/173.245.50.186|173.245.50.186]] 13:18, 19 May 2017 (UTC) sbendl

@@ Line 27: / Line 27: @@
 :Good, then maybe we won't have over-thought explanations anymore.
 :: "That was a joke, haha" [[User:Elektrizikekswerk|Elektrizikekswerk]] ([[User talk:Elektrizikekswerk|talk]]) 07:36, 18 May 2017 (UTC)
-; I lovingly think of this site as "Over-Explain XKCD"[[Special:Contributions/172.68.54.112|172.68.54.112]] 17:44, 20 May 2017 (UTC)
+: I lovingly think of this site as "Over-Explain XKCD"[[Special:Contributions/172.68.54.112|172.68.54.112]] 17:44, 20 May 2017 (UTC)
 The fuck is "{{w|Pinball}}"? [[Special:Contributions/162.158.122.66|162.158.122.66]] 03:59, 19 May 2017 (UTC)