Local and distributed concept representation via colimits: An example

Abstract An issue in the representation by a neural network of objects detected by its sensors is whether such representations are based at a single neural cell or require a distributed subnetwork. We perform an analysis near the sensor level of a simple geometric shape and show that given the appropriate mathematical structure, representations can be both local and distributed. Moreover, as simulations demonstrate, using the proposed structures in neural network design leads to improved performance in object discrimination.

[1]  A. Damasio.  Time-locked multiregional retroactivation: A systems-level proposal for the neural substrates of recall and recognition , 1989, Cognition.

[2]  S. Lane.  Categories for the Working Mathematician , 1971 .

[3]  T. P. Caudell,et al.  Aphasic compressed representations: a functorial semantic design principle for coupled ART networks , 2002, Proceedings of the 2002 International Joint Conference on Neural Networks. IJCNN'02 (Cat. No.02CH37290).

[4]  L. Barsalou,et al.  Whither structured representation? , 1999, Behavioral and Brain Sciences.

[5]  D. Amaral,et al.  Hippocampal‐neocortical interaction: A hierarchy of associativity , 2000, Hippocampus.

[6]  C. Koch,et al.  Latency and Selectivity of Single Neurons Indicate Hierarchical Processing in the Human Medial Temporal Lobe , 2008, The Journal of Neuroscience.

[7]  A. Treves,et al.  The representational capacity of the distributed encoding of information provided by populations of neurons in primate temporal visual cortex , 1997, Experimental Brain Research.

[8]  Jeffrey S Bowers,et al.  Challenging the widespread assumption that connectionism and distributed representations go hand-in-hand , 2002, Cognitive Psychology.

[9]  Jack L. Gallant,et al.  Natural Scene Statistics Account for the Representation of Scene Categories in Human Visual Cortex , 2013, Neuron.

[10]  Thomas P. Caudell,et al.  Temporal semantics: An extended definition for neural morphisms , 2009, 2009 International Joint Conference on Neural Networks.

[11]  Thomas P. Caudell,et al.  Ontologies and Worlds in Category Theory: Implications for Neural Systems , 2006 .

[12]  H. Eichenbaum,et al.  Hippocampal Neurons Encode Information about Different Types of Memory Episodes Occurring in the Same Location , 2000, Neuron.

[13]  Shawn E. Taylor,et al.  Applying category theory to improve the performance of a neural architecture , 2009, Neurocomputing.

[14]  Asim Roy,et al.  A theory of the brain: localist representation is used widely in the brain , 2012, Front. Psychology.

[15]  C. Koch,et al.  Explicit Encoding of Multimodal Percepts by Single Neurons in the Human Brain , 2009, Current Biology.

[16]  J. Bowers.  Grandmother cells and localist representations: a review of current thinking , 2017 .

[17]  Michael Barr,et al.  Category theory for computing science , 1995, Prentice Hall International Series in Computer Science.

[18]  Xueqi Cheng,et al.  Text Matching as Image Recognition , 2016, AAAI.

[19]  Yunhai Xiao,et al.  From categorical semantics to neural network design , 2003, Proceedings of the International Joint Conference on Neural Networks, 2003..

[20]  Jack L. Gallant,et al.  A Continuous Semantic Space Describes the Representation of Thousands of Object and Action Categories across the Human Brain , 2012, Neuron.

[21]  Nick Craswell,et al.  Learning to Match using Local and Distributed Representations of Text for Web Search , 2016, WWW.