Timeline for Do machine learning algorithms have knowledge (if not justified true beliefs)?
Current License: CC BY-SA 4.0
20 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Dec 3, 2019 at 10:38 | history | edited | Conifold | CC BY-SA 4.0 |
edited title
|
| Sep 24, 2018 at 15:05 | answer | added | drvrm | timeline score: 0 | |
| Oct 2, 2016 at 2:31 | history | edited | Conifold |
edited tags
|
|
| May 12, 2016 at 8:51 | history | edited | Conifold |
edited tags
|
|
| Apr 27, 2016 at 2:56 | history | edited | Conifold |
edited tags
|
|
| Mar 27, 2016 at 15:45 | comment | added | Dave | @virmaior feature selection (machinelearningmastery.com/an-introduction-to-feature-selection) is an even more relevant analog to "noticing" especially when it is executed "on-line" in repsonse to a particular test case (aclweb.org/anthology/P15-1015). I'm inclined to agree that these types of algs. are probably not models or simulations of biological noticing; but in the cases where they are used, they apply, in the terms of the overall framework of learning, a similar role to noticing. | |
| Mar 2, 2016 at 0:54 | answer | added | alanf | timeline score: 1 | |
| Jan 27, 2016 at 18:14 | vote | accept | Dave | ||
| Jan 27, 2016 at 18:14 | vote | accept | Dave | ||
| Jan 27, 2016 at 18:14 | |||||
| Jan 27, 2016 at 18:13 | vote | accept | Dave | ||
| Jan 27, 2016 at 18:14 | |||||
| Sep 25, 2015 at 8:09 | comment | added | virmaior | I'm not sure I follow you on the use of "notice" there. Noticing to me seems to be a faculty that machines lack. Having a hard-coded detection sensor seems to be completely the opposite of at least one normal sense of noticing. And machines seem (and my knowledge here might be limited) to merely learn to filter out detection information that doesn't correlate. We seem (at least I think) to do the opposite at least on the level of consciousness and call this "noticing" | |
| Sep 24, 2015 at 12:25 | vote | accept | Dave | ||
| Jan 27, 2016 at 18:13 | |||||
| Sep 24, 2015 at 12:02 | comment | added | Dave | @virmaior I interpret the fact that at the end of training the resulting model structures encode which features of the problem domain are relevant for the task at had as "noticing which features are relevant"; at least in an analogical sense. | |
| Sep 24, 2015 at 6:40 | history | tweeted | twitter.com/#!/StackPhilosophy/status/646937141755420673 | ||
| Sep 24, 2015 at 0:26 | comment | added | virmaior | There seems to be a potential disanalogy in your aside about bird watchers insofar as the meaning of "notice" when applied to a human subject might be dissimilar to anything machine learning algorithms do. Or to put it another way, the "noticing" seems to need to be programmed into a program in a way at least the programmer fundamentally understands, but birdwatchers notice in a way they may not understand. | |
| Sep 24, 2015 at 0:00 | answer | added | Conifold | timeline score: 7 | |
| Sep 23, 2015 at 16:36 | answer | added | Teagen Dix | timeline score: 3 | |
| Sep 23, 2015 at 16:18 | history | edited | Dave | CC BY-SA 3.0 |
deleted 7 characters in body
|
| Sep 23, 2015 at 13:57 | history | edited | Dave | CC BY-SA 3.0 |
added 78 characters in body
|
| Sep 23, 2015 at 13:52 | history | asked | Dave | CC BY-SA 3.0 |