國立虎尾科技大學 |

Dynamic constraints in statistical learning.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	Dynamic constraints in statistical learning./
作者:	de Leeuw, Joshua R.
面頁冊數:	1 online resource (163 pages)
附註:	Source: Dissertation Abstracts International, Volume: 78-02(E), Section: B.
標題:	Cognitive psychology. -
電子資源:	click for full text (PQDT)
ISBN:	9781339994031

Dynamic constraints in statistical learning.
de Leeuw, Joshua R.

Dynamic constraints in statistical learning. - 1 online resource (163 pages)

Source: Dissertation Abstracts International, Volume: 78-02(E), Section: B.

Thesis (Ph.D.)--Indiana University, 2016.

Includes bibliographical references

Statistical learning is a ubiquitous cognitive phenomenon in which learners extract the probabilistic regularities that generate the sensory environment. Characterizing the mechanisms that enable this kind of learning requires describing the constraints that shape learning. In this dissertation, I describe how some constraints on statistical learning may change on very short timescales as the learner acquires new information. In a series of experiments, I show that learning part of the probabilistic structure of a sequential pattern substantially improves learning of the other (statistically independent) parts of the probabilistic structure. This suggests that the learning process alters its own constraints. Furthermore, I demonstrate that the learning curves for individual items show rapid step-like changes as learners discover the statistical structure of the sequence. Taken together, these results suggest limits on the kinds of computational mechanisms that can explain statistical learning. A successful computational account needs to capture both the dependence between learning different parts of the probabilistic structure and the sudden, rather than gradual, changes in behavioral evidence of learning. I propose a general accumulator model to handle cases where people are learning multiple items simultaneously with or without mutual dependence between the learning rates. The model predicts when people will show the step-like change in behavior as they learn a sequential structure. The model is able to capture the results of the experiments only with high dependence between the learning rates of different items.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2018

Mode of access: World Wide Web

ISBN: 9781339994031Subjects--Topical Terms:

556029
Cognitive psychology.
Index Terms--Genre/Form:

554714
Electronic books.

Dynamic constraints in statistical learning.
LDR:02787ntm a2200325K 4500 001 913650
005 20180622095235.5
006 m o u
007 cr mn||||a|a||
008 190606s2016 xx obm 000 0 eng d
020 $a 9781339994031
035 $a (MiAaPQ)AAI10144241
035 $a (MiAaPQ)indiana:14309
035 $a AAI10144241
040 $a MiAaPQ $b eng $c MiAaPQ
100 1 $a de Leeuw, Joshua R. $3 1186574
245 1 0 $a Dynamic constraints in statistical learning.
264 0 $c 2016
300 $a 1 online resource (163 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertation Abstracts International, Volume: 78-02(E), Section: B.
500 $a Advisers: Robert L. Goldstone; Peter M. Todd.
502 $a Thesis (Ph.D.)--Indiana University, 2016.
504 $a Includes bibliographical references
520 $a Statistical learning is a ubiquitous cognitive phenomenon in which learners extract the probabilistic regularities that generate the sensory environment. Characterizing the mechanisms that enable this kind of learning requires describing the constraints that shape learning. In this dissertation, I describe how some constraints on statistical learning may change on very short timescales as the learner acquires new information. In a series of experiments, I show that learning part of the probabilistic structure of a sequential pattern substantially improves learning of the other (statistically independent) parts of the probabilistic structure. This suggests that the learning process alters its own constraints. Furthermore, I demonstrate that the learning curves for individual items show rapid step-like changes as learners discover the statistical structure of the sequence. Taken together, these results suggest limits on the kinds of computational mechanisms that can explain statistical learning. A successful computational account needs to capture both the dependence between learning different parts of the probabilistic structure and the sudden, rather than gradual, changes in behavioral evidence of learning. I propose a general accumulator model to handle cases where people are learning multiple items simultaneously with or without mutual dependence between the learning rates. The model predicts when people will show the step-like change in behavior as they learn a sequential structure. The model is able to capture the results of the experiments only with high dependence between the learning rates of different items.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2018
538 $a Mode of access: World Wide Web
650 4 $a Cognitive psychology. $3 556029
650 4 $a Statistics. $3 556824
655 7 $a Electronic books. $2 local $3 554714
690 $a 0633
690 $a 0463
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a Indiana University. $b Cognitive Science. $3 1186575
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10144241 $z click for full text (PQDT)