Combining causal and similarity-based reasoning

Kemp, Charles; Shafto, Patrick; Berke, Allison P.; Tenenbaum, Joshua B.

doi:10.1037/e527352012-450

Cited by 13 publications

(34 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Kemp et al (2007) have also presented a Bayesian model of inductive reasoning that incorporates both knowledge about relationships between objects and knowledge about relationships between object properties. The model is useful to infer some properties of previously unseen stimuli, based on the learned relationships between objects.…”

Section: Contents Lists Available At Sciencedirectmentioning

confidence: 99%

Quantitative modeling of the neural representation of objects: How semantic feature norms can account for fMRI activation

2011

View full text Add to dashboard Cite

Recent multivariate analyses of fMRI activation have shown that discriminative classifiers such as Support Vector Machines (SVM) are capable of decoding fMRI-sensed neural states associated with the visual presentation of categories of various objects. However, the lack of a generative model of neural activity limits the generality of these discriminative classifiers for understanding the underlying neural representation. In this study, we propose a generative classifier that models the hidden factors that underpin the neural representation of objects, using a multivariate multiple linear regression model. The results indicate that object features derived from an independent behavioral feature norming study can explain a significant portion of the systematic variance in the neural activity observed in an object-contemplation task. Furthermore, the resulting regression model is useful for classifying a previously unseen neural activation vector, indicating that the distributed pattern of neural activities encodes sufficient signal to discriminate differences among stimuli. More importantly, there appears to be a double dissociation between the two classifier approaches and within-versus between-participants generalization. Whereas an SVM-based discriminative classifier achieves the best classification accuracy in within-participants analysis, the generative classifier outperforms an SVM-based model which does not utilize such intermediate representations in between-participants analysis. This pattern of results suggests the SVM-based classifier may be picking up some idiosyncratic patterns that do not generalize well across participants and that good generalization across participants may require broad, large-scale patterns that are used in our set of intermediate semantic features. Finally, this intermediate representation allows us to extrapolate the model of the neural activity to previously unseen words, which cannot be done with a discriminative classifier.

show abstract

Section: Contents Lists Available At Sciencedirectmentioning

confidence: 99%

Quantitative modeling of the neural representation of objects: How semantic feature norms can account for fMRI activation

2011

View full text Add to dashboard Cite

show abstract

“…Human similarity judgments are sometimes asymmetric (Tversky, 1977; but see Aguilar & Medin, 1999), and human inductive judgment is often incoherent (Kahneman & Tversky, 1983;Tentori et al, 2004) and informed by knowledge beyond similarity, such as causal knowledge and logical knowledge (Kemp, Shafto, Berke & Tenenbaum, 2007). Nonetheless, the present work provides evidence that, at least in propitious circumstances, similarity in the structure of brain activity can underwrite inductive judgment.…”

Section: Resultsmentioning

confidence: 66%

Category-based induction from similarity of neural activation

Weber

Osherson

2013

Cogn Affect Behav Neurosci

View full text Add to dashboard Cite

“…Therefore, it would be reasonable to predict that this new bird is unlikely to fly. (For a similar example about predicting lung cancer, see Kemp, Shafto, Berke & Tenenbaum, 2007).…”

Section: Informants Of Feature Inferencementioning

confidence: 99%

Proceedings of the 9th Conference of the Australasian Society for Cognitive Science

2010

View full text Add to dashboard Cite

. Submissions to the conference were reviewed by the Program Committee, who came up with an excellent program of 106 presented papers (in three parallel-track sessions, and including eight submitted symposia) and 20 posters (on show throughout the conference, but highlighted in an enjoyable poster session over drinks on day 2). The full program, along with other information about the conference, remains online at http://www.maccs.mq.edu.au/ascs09. The Australasian Society for Cognitive Science has had an intermittent formal existence, driven mainly by the organization of its string of enjoyable, successful, and highly sociable conferences, which have taken place roughly every two years since the inaugural conference, organized by Peter Slezak, at the University of New South Wales in 1990 (a full list of past conferences is appended below).At the 9 th conference at Macquarie, 163 people attended, including substantial numbers from overseas and interstate, enjoying both the sweet Sydney spring weather and three stimulating days of talks and talk. There was no specific theme, but in addition to specialist research papers within specific sub/disciplines, we encouraged integrative papers which forge connections between sub/disciplines. This integrative ideal also informed invitations to our five excellent keynote speakers: Stephen Crain (Macquarie Centre for Cognitive Science), Jakob Hohwy (Philosophy, Monash University), Jason Mattingley (Psychology & Queensland Brain Institute, University of Queensland), Thomas Metzinger (Philosophy, Johannes Gutenberg-Universität, Mainz; & Institute for Advanced Study, Berlin, Germany, and Barbara Tversky (Psychology, Stanford University; & Columbia University, USA). Versions of three of these keynote talks appear in these proceedings: the papers by Jason Mattingley and Thomas Metzinger have been published elsewhere.Perhaps because the cognitive science community in Australasia is relatively small, compared at least with our counterparts in Asia, Europe, and North America, it retains a thoroughly multidisciplinary atmosphere. The programme confirmed the striking pluralism of the contemporary cognitive sciences, drawing on many distinctive disciplines as researchers on cognition are looking both down into the brain and out into the environment. Most contributions were from psychologists, with sizable numbers from philosophy and linguistics: but many other areas were also represented, including computational neuroscience, neuropsychiatry, clinical psychology, computer science, human-computer interaction, anthropology, education, history and philosophy of science, cognitive archaeology, cognitive ethology, English literature, and cultural history.All those who had presented papers and posters were invited to submit revised versions, following the conference, as full contributions to the published Proceedings. All submissions were reviewed afresh, and revisions were submitted in early 2010: we are delighted to thank our team of reviewers for their hard and effective work.Fifty-seven p...

show abstract

Combining causal and similarity-based reasoning

Cited by 13 publications

References 10 publications

Quantitative modeling of the neural representation of objects: How semantic feature norms can account for fMRI activation

Quantitative modeling of the neural representation of objects: How semantic feature norms can account for fMRI activation

Category-based induction from similarity of neural activation

Proceedings of the 9th Conference of the Australasian Society for Cognitive Science

Contact Info

Product

Resources

About