The Simplicity Principle in Human Concept Learning

@article{Feldman2003TheSP,
  title={The Simplicity Principle in Human Concept Learning},
  author={Jacob Feldman},
  journal={Current Directions in Psychological Science},
  year={2003},
  volume={12},
  pages={227 - 232}
}
  • J. Feldman
  • Published 2003
  • Psychology
  • Current Directions in Psychological Science
How do we learn concepts and categories from examples? Part of the answer might be that we induce the simplest category consistent with a given set of example objects. This seemingly obvious idea, akin to simplicity principles in many fields, plays surprisingly little role in contemporary theories of concept learning, which are mostly based on the storage of exemplars, and avoid summarization or overt abstraction of any kind. This article reviews some evidence that complexity minimization does… Expand
A logical framework to study concept-learning biases in the presence of multiple explanations.
TLDR
This work exploits the tools of propositional logic to develop an experimental framework that fully controls which minimal rules compete to explain the examples and is able to recover the features used by the participant to build the classification rule, without relying on supplementary attention-tracking mechanisms. Expand
Finding categories through words: More nameable features improve category learning
TLDR
It is suggested that compact verbal labels may facilitate hypothesis formation during learning: it is easier to pose the hypothesis "it is about redness" than " it is about that pinkish-purplish color". Expand
Classifying partial exemplars: seeing less and learning more.
  • E. Taylor, B. Ross
  • Psychology, Medicine
  • Journal of experimental psychology. Learning, memory, and cognition
  • 2009
TLDR
In 2 experiments, the authors compared standard classification learning with a slightly different task where items appeared with occluded features, offering evidence that classification can lead to learning within-category information. Expand
In search of abstraction: The varying abstraction model of categorization
TLDR
The varying abstraction model of categorization (VAM) is proposed, which investigates the possibility that partial abstraction may play a role in category learning and provides support for the idea that some form of partial abstraction can be used in people’s category representations. Expand
A Rational Analysis of Rule-Based Concept Learning
TLDR
A new model of human concept learning that provides a rational analysis of learning feature-based concepts is proposed, built upon Bayesian inference for a grammatically structured hypothesis space-a concept language of logical rules. Expand
Complexity minimization in rule-based category learning: Revising the catalog of Boolean concepts and evidence for non-minimal rules
TLDR
It is demonstrated that the minimum algebraic description is very useful for describing the most economical categorization rule that can be used, but does not always correspond to the category definition used by a participant in a concept-learning task. Expand
Towards a more flexible language of thought: Bayesian grammar updates after each concept exposure.
TLDR
This work performs a concept learning experiment that demonstrates that humans can change very rapidly the repertoire of symbols they use to identify concepts, by compiling expressions that are frequently used into new symbols of the language. Expand
A Developmental Approach to Concept Learning
TLDR
The proposed system learns abstract, generic features of objects and associates words from sentences referring to those objects with the features, thus providing a grounding for the meaning of the words, and allows the system to later identify such features in previously unseen images. Expand
Abstraction and model evaluation in category learning
TLDR
It is demonstrated that, across the 30 data sets, complexity differences between the models in the VAM family are small, indicating that a (computationally challenging) complexity-sensitive model evaluation method is uncalled for, and that the use of a complexity-insensitive model Evaluation method is justified. Expand
Learning and the language of thought
This thesis develops the hypothesis that key aspects of learning and development can be understood as rational statistical inferences over a compositionally structured representation system, aExpand
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 53 REFERENCES
Minimization of Boolean complexity in human concept learning
TLDR
The data reveal a surprisingly simple empirical ‘law’: the subjective difficulty of a concept is directly proportional to its Boolean complexity (the length of the shortest logically equivalent propositional formula)—that is, to its logical incompressibility. Expand
Categorization Under Complexity: A Unified MDL Account of Human Learning of Regular and Irregular Categories
TLDR
It is suggested that the intrinsic complexity of a concept (that is, its description length) systematically influences its learnability. Expand
The Structure of Perceptual Categories
  • Feldman
  • Mathematics, Medicine
  • Journal of mathematical psychology
  • 1997
TLDR
The Genericity Constraint dictates that among all the models on the lattice that apply, the observer should choose the one in which the observed object is generic, which is simply the lowest in the partial order. Expand
Thirty categorization results in search of a model.
  • J. Smith, J. P. Minda
  • Psychology, Medicine
  • Journal of experimental psychology. Learning, memory, and cognition
  • 2000
TLDR
The authors conclude that the data produced by this category structure do not clearly support exemplar theory, and a simpler exemplar process than assumed by exemplar models suffices to explain these data sets. Expand
Similarity and rules: distinct? exhaustive? empirically distinguishable?
TLDR
This work proposes a 'core' distinction between rule- and similarity-based processes, in terms of the way representations of stored information are 'matched' with the representation of a novel item, and argues that the AI literature suggests that approaches which combine rules and similarity are an important new focus for empirical work. Expand
Simplicity: a unifying principle in cognitive science?
TLDR
Research exploring the idea that simplicity drives a wide range of cognitive processes is reviewed, outlining mathematical theory, computational results and empirical data that underpin this viewpoint. Expand
ALCOVE: an exemplar-based connectionist model of category learning.
  • J. Kruschke
  • Psychology, Medicine
  • Psychological review
  • 1992
TLDR
Alcove selectively attends to relevant stimulus dimensions, can account for a form of base-rate neglect, does not suffer catastrophic forgetting, and can exhibit 3-stage learning of high-frequency exceptions to rules, whereas such effects are not easily accounted for by models using other combinations of representation and learning method. Expand
Exemplar-Based Accounts of Relations Between Classification, Recognition, and Typicality
Previously published sets of classification and old-new recognition memory data are reanalyzed within the framework of an exemplar-based generalization model. The key assumption in the model is that,Expand
Constraints and Preferences in Inductive Learning: An Experimental Study of Human and Machine Performance
TLDR
Al induction programs become condidote psychological process models ond results from inductive learning experiments con suggest new algorithms, and results show that humon inductive generolizotions tend toword greater specificity than would be expected if conceptual simplicity were the key constraint on inductions. Expand
A catalog of Boolean concepts
Boolean concepts are concepts whose membership is determined by a Boolean function, such as that expressed by a formula of propositional logic. Certain Boolean concepts have been much studied in theExpand
...
1
2
3
4
5
...