FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2772430276> ?p ?o ?g. }

• W2772430276 abstract "Error-Reduction and Simplicity: Opposing Goals in Classification Learning Mark Blair (mrblair@indiana.edu) Indiana University, Department of Psychology 1101 E. Tenth Street, Bloomington, IN 47405 USA Abstract which the categories separate, can be expensive to represent. Completely altering the dimensions used for categorization can require more energy than using unique stimulus-level elements to classify exceptions. This leads to the memorization of exceptions, rather than a refining of the cue set. It is clear that category learning is influenced by opposing forces; one to enlarge, and one to reduce the dimensionality of the cue set. The present research examines the complementary processes of the expansion and reduction of the cue set toward effective and efficient representation. The goal of the present study is to verify that participants can and do optimize their cue sets using both expansion and reduction when learning categories with subtle dimensions. Previous research has demonstrated these complimentary processes by manipulating the stimulus set to provide a new dimension (Blair & Homa, 2003b). These studies produced wide individual differences, with many participants incorporating new dimensions to eliminate category exceptions, and many others choosing to rely on simple cue sets which result in many exceptions and significant error. The Blair and Homa (2003b) studies also showed that participants can shift to optimal spaces if they are less complex. Overall, in these studies participants demonstrated both the flexibility found in studies of expertise and the insistence on simplicity found in studies of knowledge restructuring. Many real world category learning problems do not involve learning new information never experienced before, but rather involve learning to be sensitive to stimulus features that have existed all along, but may have been overlooked for more salient dimensions. For example in bird identification, color is an obvious perceptual cue, used by experts and novices alike. To tell the difference between a Hepatic Tanager and a Summer Tanager, both of which are predominantly red, one must notice the color of the bill and whether the bird has a gray ear patch or not. These are features that novices are prone to miss. The present studies use two obvious dimensions as well as a subtle third dimension as a direct analogy to those cases. Studies of real world experts show that they use different and subtler regularities than novices to make effective classifications. In laboratory studies of learning however, participants have a strong preference for simple cue sets, even at the expense of accuracy. The present experiments investigate participants’ ability to use subtle stimulus dimensions in order to eliminate category exceptions. Results show that some participants were able to use the optimal 3- cue set, but many could not. When there were two optimal cue sets, one with 2 dimensions and one with 3, participants favored the simpler set, even though it meant ignoring an obvious and diagnostic cue. Overall there were wide individual differences, with almost every cue set adopted by some participants. Current theories of attention that posit rapid shifts of learned attention offer promise in accounting for the results. For any organism to adapt successfully it must become sensitive to meaningful regularities in the environment. Humans have developed flexible learning systems, allowing them to rapidly adjust to changing environments. Learning concepts, representations of classes of stimuli that 4require an equivalent response, conserves resources by reducing the amount of information that needs to be processed from the environment, and also allows for generalization to related, novel circumstances. Representing a complex environment, with abundant interdependencies and subtle regularities requires a rich set of concepts. To perform this function, the human perceptual system can attend selectively, become sensitized to highly complex stimulus dimensions, and even create novel functional features. These processes affect the perception of a stimulus and therefore alter the representation of that stimulus. Differences in cue use and representation across experts and novices appear in many areas such as biology (Boster & Johnson, 1989), physics (Chi, Feltovich, & Glaser, 1981), computer programming (Davies, 1994), wine tasting (Solomon, 1997), bird identification (Johnson & Mervis, 1997) and pocket billiards (Blair & McBeath, 2001). Despite an abundance of differences between expert and novice differences in cue use, laboratory studies of learning (specifically “knowledge restructuring”) have shown that participants have had a strong resistance to using new information and more complicated cue sets, even though they would afford better performance (Lewandowsky, Kalish & Griffiths, 2000). To the extent that learning is error driven, exceptions in the cue set provide a powerful motivator to incorporate new dimensions, however, there is also a pressure toward simplicity. Additional dimensions, which may a space in Experiment 1 In Experiment 1, a sequential presentation same-different task was employed to examine the discriminability of the three dimensions used in the remaining experiments, and an additional dimension (color) used in a related set of studies (Blair & Homa, 2003b). If the stimuli are to be used in later experiments, they should be of roughly equal discriminability, with the exception of tail bumpiness, which should be significantly less discriminable than the other three dimensions." @default.
• W2772430276 created "2017-12-22" @default.
• W2772430276 creator A5008458432 @default.
• W2772430276 date "2004-01-01" @default.
• W2772430276 modified "2023-09-24" @default.
• W2772430276 title "Error-Reduction and Simplicity: Opposing Goals in Classification Learning" @default.
• W2772430276 cites W2000405686 @default.
• W2772430276 cites W2040848903 @default.
• W2772430276 cites W2045300586 @default.
• W2772430276 cites W2045466271 @default.
• W2772430276 cites W2057537725 @default.
• W2772430276 cites W2060535027 @default.
• W2772430276 cites W2075249440 @default.
• W2772430276 cites W2077421553 @default.
• W2772430276 cites W2094583869 @default.
• W2772430276 cites W2115693236 @default.
• W2772430276 cites W2128926256 @default.
• W2772430276 cites W2168729764 @default.
• W2772430276 cites W2169533403 @default.
• W2772430276 hasPublicationYear "2004" @default.
• W2772430276 type Work @default.
• W2772430276 sameAs 2772430276 @default.
• W2772430276 citedByCount "0" @default.
• W2772430276 crossrefType "journal-article" @default.
• W2772430276 hasAuthorship W2772430276A5008458432 @default.
• W2772430276 hasConcept C105795698 @default.
• W2772430276 hasConcept C111472728 @default.
• W2772430276 hasConcept C119857082 @default.
• W2772430276 hasConcept C138885662 @default.
• W2772430276 hasConcept C154945302 @default.
• W2772430276 hasConcept C15744967 @default.
• W2772430276 hasConcept C169760540 @default.
• W2772430276 hasConcept C169900460 @default.
• W2772430276 hasConcept C177264268 @default.
• W2772430276 hasConcept C180747234 @default.
• W2772430276 hasConcept C199360897 @default.
• W2772430276 hasConcept C2776372474 @default.
• W2772430276 hasConcept C2780598303 @default.
• W2772430276 hasConcept C33923547 @default.
• W2772430276 hasConcept C41008148 @default.
• W2772430276 hasConcept C66102695 @default.
• W2772430276 hasConcept C70518039 @default.
• W2772430276 hasConcept C94124525 @default.
• W2772430276 hasConceptScore W2772430276C105795698 @default.
• W2772430276 hasConceptScore W2772430276C111472728 @default.
• W2772430276 hasConceptScore W2772430276C119857082 @default.
• W2772430276 hasConceptScore W2772430276C138885662 @default.
• W2772430276 hasConceptScore W2772430276C154945302 @default.
• W2772430276 hasConceptScore W2772430276C15744967 @default.
• W2772430276 hasConceptScore W2772430276C169760540 @default.
• W2772430276 hasConceptScore W2772430276C169900460 @default.
• W2772430276 hasConceptScore W2772430276C177264268 @default.
• W2772430276 hasConceptScore W2772430276C180747234 @default.
• W2772430276 hasConceptScore W2772430276C199360897 @default.
• W2772430276 hasConceptScore W2772430276C2776372474 @default.
• W2772430276 hasConceptScore W2772430276C2780598303 @default.
• W2772430276 hasConceptScore W2772430276C33923547 @default.
• W2772430276 hasConceptScore W2772430276C41008148 @default.
• W2772430276 hasConceptScore W2772430276C66102695 @default.
• W2772430276 hasConceptScore W2772430276C70518039 @default.
• W2772430276 hasConceptScore W2772430276C94124525 @default.
• W2772430276 hasIssue "26" @default.
• W2772430276 hasLocation W27724302761 @default.
• W2772430276 hasOpenAccess W2772430276 @default.
• W2772430276 hasPrimaryLocation W27724302761 @default.
• W2772430276 hasRelatedWork W117379691 @default.
• W2772430276 hasRelatedWork W1489568664 @default.
• W2772430276 hasRelatedWork W1543409941 @default.
• W2772430276 hasRelatedWork W1975861865 @default.
• W2772430276 hasRelatedWork W199069442 @default.
• W2772430276 hasRelatedWork W1999297315 @default.
• W2772430276 hasRelatedWork W2063950359 @default.
• W2772430276 hasRelatedWork W2183941720 @default.
• W2772430276 hasRelatedWork W2222157309 @default.
• W2772430276 hasRelatedWork W2396937662 @default.
• W2772430276 hasRelatedWork W2408703897 @default.
• W2772430276 hasRelatedWork W2417128737 @default.
• W2772430276 hasRelatedWork W2573837299 @default.
• W2772430276 hasRelatedWork W2944981745 @default.
• W2772430276 hasRelatedWork W3045554139 @default.
• W2772430276 hasRelatedWork W339150178 @default.
• W2772430276 hasRelatedWork W72584535 @default.
• W2772430276 hasRelatedWork W144090452 @default.
• W2772430276 hasRelatedWork W2185019812 @default.
• W2772430276 hasRelatedWork W2339899919 @default.
• W2772430276 hasVolume "26" @default.
• W2772430276 isParatext "false" @default.
• W2772430276 isRetracted "false" @default.
• W2772430276 magId "2772430276" @default.
• W2772430276 workType "article" @default.