Filters








128 Hits in 2.3 sec

Challenges for interactivist-constructivist robotics

Răzvan V. Florian
2010 New Ideas in Psychology  
A control system that could support this kind of learning should be a collection of parallel, heterogenous, loosely coupled processes, capable of self-organization, such as a neural network (Pfeifer and  ...  a sound framework for understanding cognition and representation and for designing genuinely intelligent artificial systems.  ...  Spike-timing-dependent plasticity (STDP) is a type of neural plasticity where synaptic changes depend on the relative timing of pre-and postsynaptic action potentials (Markram et al., 1997; Bi and Poo  ... 
doi:10.1016/j.newideapsych.2009.09.009 fatcat:qx7sd2lyszcr5gupgdcwq5q5xy

A Cognitive Architecture Based on a Learning Classifier System with Spiking Classifiers [article]

David Howard, Larry Bull, Pier-Luca Lanzi
2015 arXiv   pre-print
The combination of temporal reinforcement learning and neural information processing is shown to outperform benchmark neural classifier systems, and successfully solve a robotic navigation task.  ...  The spiking classifiers are coupled with a temporally-sensitive reinforcement learning algorithm, which allows the system to perform temporal state decomposition by appropriately rewarding "macro-actions  ...  Neural Representations in Learning Classifier Systems Neural networks have been used in LCS to compute predictions [31] , and as a direct replacement for classifiers.  ... 
arXiv:1508.07700v1 fatcat:7loktdgxybe3bmc7sgnpcvjvzm

A Spiking Neural Learning Classifier System [article]

Gerard Howard and Larry Bull and Pier-Luca Lanzi
2012 arXiv   pre-print
Learning Classifier Systems (LCS) are population-based reinforcement learners used in a wide variety of applications.  ...  We employ a constructivist model of growth of both neurons and dendrites that realise flexible learning by evolving structures of sufficient complexity to solve a well-known problem involving continuous  ...  Our specific approach to temporal machine learning involves the use of a Learning Classifier System, or LCS (Holland, 1976) , a form of online evolutionary reinforcement-based learning that evolves a  ... 
arXiv:1201.3249v1 fatcat:kimwxfr2vjc5ncmjwyw5wbr4uq

Study of Transmission Line Boundary Protection Using a Multilayer Perceptron Neural Network with Back Propagation and Wavelet Transform

Daniel Okojie, Linus Idoko, Daniel Herbert, Agha Nnachi
2021 Applied System Innovation  
These components are then used as the input data for a multilayer perceptron neural network with backpropagation that can classify between different fault locations in the system.  ...  The outcome also shows that more accurate fault identification and classification are achievable by using neural network than by the conventional system currently in use.  ...  Unfortunately, these lines are not perfect and are susceptible to faults that can cause high voltage and current spikes within the system.  ... 
doi:10.3390/asi4040095 fatcat:l46vfy2bgvehbeidjuoalsjulu

The neural basis of cognitive development: a constructivist manifesto

S R Quartz, T J Sejnowski
1997 Behavioral and Brain Sciences  
This uniquely powerful and general learning strategy undermines the central assumption of classical learnability theory, that the learning properties of a system can be deduced from a fixed computational  ...  The interaction between the environment and neural growth results in a flexible type of learning: "constructive learning" minimizes the need for prespecification in accordance with recent neurobiological  ...  The learning of grammars is a different and harder problem even for constructivist neural networks.  ... 
pmid:10097006 fatcat:ammov422u5dy7aouvbm5ezhg2e

The neural basis of cognitive development: A constructivist manifesto

Steven R. Quartz, Terrence J. Sejnowski
1997 Behavioral and Brain Sciences  
This uniquely powerful and general learning strategy undermines the central assumption of classical learnability theory, that the learning properties of a system can be deduced from a fixed computational  ...  The interaction between the environment and neural growth results in a flexible type of learning: "constructive learning" minimizes the need for prespecification in accordance with recent neurobiological  ...  The learning of grammars is a different and harder problem even for constructivist neural networks.  ... 
doi:10.1017/s0140525x97001581 fatcat:2x57otwpxrcwpklsixe3sukshu

Artificial Neurogenesis: An Introduction and Selective Review [chapter]

Taras Kowaliw, Nicolas Bredeche, Sylvain Chevallier, René Doursat
2014 Studies in Computational Intelligence  
This has important consequences on any attempt to classify and model neural types [268] .  ...  Both selectivist and constructivist theories of brain development posit a central role for environmental stimuli in the generation of neural morphology.  ... 
doi:10.1007/978-3-642-55337-0_1 fatcat:xx6nzfvbmfgzjhse6t5il3lbxe

Making AI 'Smart': Bridging AI and Cognitive Science [article]

Madhav Agarwal, Siddhant Bansal
2022 arXiv   pre-print
This will help develop more powerful AI systems and simultaneously gives us a better understanding of how the human brain works.  ...  We discuss the various possibilities and challenges of bridging these two fields and how they can benefit each other.  ...  They showed that spiking neural network trained with spike-timing-dependent plasticity gives better generalization and robustness on novel inputs when an offline, sleep-like period is used after training  ... 
arXiv:2112.15360v2 fatcat:p257psawpbhiferqrufftdy7aq

Big data and the industrialization of neuroscience: A safe roadmap for understanding the brain?

Yves Frégnac
2017 Science  
In view of the weight presently given to spikebased feedforward processing and deep learning, the reexamination of conductance-based versus spike-based computing and the role given to synaptic reentry  ...  By applying unsupervised learning methods to the largest possible set of coregistered neural data and behavioral observations, one may hope to achieve substantial dimensionality reduction and obtain an  ... 
doi:10.1126/science.aan8866 pmid:29074766 fatcat:7wo22kixfvdojoecsrmiao32zq

Unsupervised Neural Network Models of the Ventral Visual Stream [article]

Chengxu Zhuang, Siming Yan, Aran Nayebi, Martin Schrimpf, Michael Frank, James DiCarlo, Daniel Yamins
2020 bioRxiv   pre-print
We find that neural network models learned with deep unsupervised contrastive embedding methods achieve neural prediction accuracy in multiple ventral visual cortical areas that equals or exceeds that  ...  of models derived using today's best supervised methods, and that the mapping of these neural network models' hidden layers is neuroanatomically consistent across the ventral stream.  ...  Neural system identification for large 428 428 populations separating what and where.  ... 
doi:10.1101/2020.06.16.155556 fatcat:qdya2vcpxveyrpxeuso6zy56ju

Computational physics of the mind

Włodzisław Duch
1996 Computer Physics Communications  
Computational physics allows to simulate complex neural processes giving a chance to answer not only the original psychophysical questions but also to create models of mind.  ...  In the XIX century and earlier such physicists as Newton, Mayer, Hooke, Helmholtz and Mach were actively engaged in the research on psychophysics, trying to relate psychological sensations to intensities  ...  Skarda and Freeman [36] have analyzed this system using biologically plausible spiking neurons.  ... 
doi:10.1016/0010-4655(96)00027-6 fatcat:ji6w2qbz7nbsze4dj5ccw2mohq

Mental Illness from the Perspective of Theoretical Neuroscience

Paul Thagard
2008 Perspectives in biology and medicine  
to solve the explanation problem of causally connecting neural processes with the behaviors and experiences found in mental illnesses.  ...  Theoretical neuroscience, which characterizes neural mechanisms using mathematical and computational models, is highly relevant to central problems in the philosophy of psychiatry.These models can help  ...  the brain and nervous system.  ... 
doi:10.1353/pbm.0.0030 pmid:18723939 fatcat:whmznpk2tvdtzinndwjyogetxa

Deep Gated Hebbian Predictive Coding Accounts for Emergence of Complex Neural Response Properties Along the Visual Cortical Hierarchy

Shirin Dora, Sander M. Bohte, Cyriel M. A. Pennartz
2021 Frontiers in Computational Neuroscience  
Here, we developed a scalable, deep network architecture for predictive coding that is trained using a gated Hebbian learning rule and mimics the feedforward and feedback connectivity of the cortex.  ...  We analyzed low- and high-level properties such as orientation selectivity, object selectivity and sparseness of neuronal populations in the model.  ...  ACKNOWLEDGMENTS We would like to thank Walter Senn and Mihai Petrovici for helpful discussions and Sandra Diaz, Anna Lührs, and Thomas Lippert for the use of supercomputers at the Jülich Supercomputing  ... 
doi:10.3389/fncom.2021.666131 fatcat:iqcqxnepavg2ra5nu4vxnsawde

Towards a computational theory of experience

Tomer Fekete, Shimon Edelman
2011 Consciousness and Cognition  
The complementary tasks of explaining what it means for a system to give rise to experience and what constitutes its content (qualia) in computational terms are exacerbated by the multiple realizability  ...  We identify conditions that a computational theory must satisfy for it to be not just a sufficient but a necessary and intrinsic explanation of qualia.  ...  We thank Kat Agres, David Chalmers, Axel Cleeremans, Rick Dale, Barbara Finlay, Rafael Malach, Björn Merker, Thomas Metzinger, Helene Porte, Michael Spivey, and two anonymous reviewers for their comments  ... 
doi:10.1016/j.concog.2011.02.010 pmid:21388834 fatcat:6ggbauvu4zb4xh546tkkiomsmy

A review of affective computing: From unimodal analysis to multimodal fusion

Soujanya Poria, Erik Cambria, Rajiv Bajpai, Amir Hussain
2017 Information Fusion  
, to cognitive and social sciences.  ...  Multimodality is defined by the presence of more than one modality or channel, e.g., visual, audio, text, gestures, and eye gage.  ...  The features were fed as one second window interval definitions, into two classifiers: SVM and a conventional Neural Network (NN).  ... 
doi:10.1016/j.inffus.2017.02.003 fatcat:ytebhjxlz5bvxcdghg4wxbvr6a
« Previous Showing results 1 — 15 out of 128 results