Filters








2,448 Hits in 5.2 sec

Introduction

1991 Machine Learning  
Pollack looks more closely at a connectionist network as a continuous dynamical system. He describes a new type of machine learning phenomenon: induction by phase transition.  ...  The papers in this special issue represent some of the best connectionist work to date on the problem of language learning.  ... 
doi:10.1007/bf00114840 fatcat:ufmubwbmczg35kg2rlr6po6a2q

:{unav)

David S. Touretzky
2012 Machine Learning  
Pollack looks more closely at a connectionist network as a continuous dynamical system. He describes a new type of machine learning phenomenon: induction by phase transition.  ...  The papers in this special issue represent some of the best connectionist work to date on the problem of language learning.  ... 
doi:10.1023/a:1022668727419 fatcat:bx3uufaz5ngvna7q6luz6ie7oa

Page 1307 of Linguistics and Language Behavior Abstracts: LLBA Vol. 26, Issue 3 [page]

1992 Linguistics and Language Behavior Abstracts: LLBA  
perspective of dynamical systems yields two interesting discoveries: (1) a longitudi- nal examination of the learning process illustrates a new form of mechan- ical inference—induction by phase transition  ...  Some practical properties of the learning algorithm are characterized. 12 Figures, 30 References. Adapted from the source document.  ... 

Letting structure emerge: connectionist and dynamical systems approaches to cognition

James L. McClelland, Matthew M. Botvinick, David C. Noelle, David C. Plaut, Timothy T. Rogers, Mark S. Seidenberg, Linda B. Smith
2010 Trends in Cognitive Sciences  
Gradient-descent learning: learning algorithms based on minimizing the error of a system (or maximizing the likelihood of the observed data) by modifying the parameters of the system based on the derivative  ...  Typical connectionist models, by contrast, follow a bottom-up approach, beginning with a characterization of neural mechanisms and exploring what macro-level functional phenomena might emerge.  ...  Gradient-descent learning: learning algorithms based on minimizing the error of a system (or maximizing the likelihood of the observed data) by modifying the parameters of the system based on the derivative  ... 
doi:10.1016/j.tics.2010.06.002 pmid:20598626 pmcid:PMC3056446 fatcat:jvoodhp3c5euhacn77yt6d2r5e

Probabilistic models of cognition: exploring representations and inductive biases

Thomas L. Griffiths, Nick Chater, Charles Kemp, Amy Perfors, Joshua B. Tenenbaum
2010 Trends in Cognitive Sciences  
Gradient-descent learning: learning algorithms based on minimizing the error of a system (or maximizing the likelihood of the observed data) by modifying the parameters of the system based on the derivative  ...  Typical connectionist models, by contrast, follow a bottom-up approach, beginning with a characterization of neural mechanisms and exploring what macro-level functional phenomena might emerge.  ...  Gradient-descent learning: learning algorithms based on minimizing the error of a system (or maximizing the likelihood of the observed data) by modifying the parameters of the system based on the derivative  ... 
doi:10.1016/j.tics.2010.05.004 pmid:20576465 fatcat:rhtsthe6pncu3dldvyf67wdpfu

Connectionism, Learning and Meaning

MORTEN H. CHRISTIANSEN, NICK CHATER
1992 Connection science  
However, we suggest that philosophy may be ill-advised to ignore the development of connectionism, particularly if connectionist systems prove to be able to learn to handle structured representations.  ...  In addition, since connectionist representations typically are ascribed content through semantic interpretation based on correlation, connectionism is prone to a number of well-known philosophical problems  ...  This position is particularly interesting in the present context, in view of the importance of learning in connectionist systems.  ... 
doi:10.1080/09540099208946617 fatcat:qsmdbbzkenb57b65mp3yiw6ete

Special issue on integration of symbolic and connectionist systems

Paolo Frasconi, Marco Gori, Franz Kurfess, Alessandro Sperduti
2002 Cognitive Systems Research  
tion to the conceptual discussion of the framework, the authors apply it to a pulsed neural network for auditory processing.  ...  This type of property is naturally expressed as a real number and thus symbolic systems have problems in dealing with this kind of tasks, while a connectionist system is often able to reach a satisfactory  ...  After formally identifying some properties required for human understandabili ty of a fuzzy knowledge base, it is argued that unconstrained learning approaches cannot guarantee that such properties will  ... 
doi:10.1016/s1389-0417(02)00057-8 fatcat:kbx3raunuzbpvpmnadmnna4sj4

Symbolic, Conceptual and Subconceptual Representations [chapter]

Peter Gärdenfors
1997 Human and Machine Perception  
For both kinds of goals, a key problem is how the information used by the cognitive system is to be modelled in an appropriate way.  ...  Again, conceptual representations should not be seen as competing with symbolic or connectionistic representations. Rather, the three kinds can be seen as three levels of  ...  Apart from the basic frequency dimension of tones, we can find some interesting further structure in the mental representation of tones.  ... 
doi:10.1007/978-1-4615-5965-8_18 fatcat:rwvmbjizijgmxmhfwhmt4oxbiy

Learning and programming in classifier systems

Richard K. Belew, Stephanie Forrest
1988 Machine Learning  
Classifier systems are low-level learning systems that are also capable of supporting representations at the symbolic level, in this paper, we explore in detail the issues surrounding the integration of  ...  programmed and learned knowledge in classifier-system representations, including comprehensibility, ease of expression, explanation, predictability, robustness, redundancy, stability, and the use of analogical  ...  We would also like to acknowledge the profound influence John Holland has had on our understanding of intelligent and adaptive systems.  ... 
doi:10.1007/bf00113897 fatcat:5e6gj7edynf37m727apu3xvkcq

HYCONES: a hybrid connectionist expert system

B de F Leão, E B Reátegui
1993 Proceedings. Symposium on Computer Applications in Medical Care  
This paper describes HYCONES, a tightly-coupled Hybrid Connectionist Expert System that integrates neural networks with a symbolic approach (frames).  ...  The symbolic paradigm provides rich and flexible constructs to describe the domain knowledge, while the connectionist one provides the system with learning capabilities.  ...  INTRODUCTION Connectionist systems attract increasing interest for their inherent learning abilities.  ... 
pmid:8130516 pmcid:PMC2248551 fatcat:3nwshyh2cbey5lqlj2gdtckz6q

Knowledge representation in machine learning [chapter]

Filippo Neri, Lorenza Saitta
1994 Lecture Notes in Computer Science  
This paper investigates the influence of knowledge representation languages on the complexity of the learning process.  ...  However, the aim of the paper is not to give a state-of-the-art account of the involved issues, but to survey the underlying ideas.  ...  Even if these approaches have been able to offer solutions to some interesting real problems, a large scale application of automatic learning techniques to real life has still to come.  ... 
doi:10.1007/3-540-57868-4_48 fatcat:xj2qfe6o7jcrnmpdevb2rnjiku

An Experimental Comparison of Symbolic and Connectionist Learning Algorithms

Raymond J. Mooney, Jude W. Shavlik, Geoffrey G. Towell, Alan Gove
1989 International Joint Conference on Artificial Intelligence  
However, the probability of correctly classifying new examples is about the same for the three systems. On noisy data sets there is some indication that backpropagation classifies more accurately.  ...  Despite the fact that many symbolic and connectionist (neural net) learning algorithms are addressing the same problem of learning from classified examples, very little Is known regarding their comparative  ...  Professor James Jerger of the Baylor College of Medicine; Rob Holte and Peter Clark for Alen Shapiro's chess data; and Terry Sejnowski for the NETtalk data.  ... 
dblp:conf/ijcai/MooneySTG89 fatcat:y6by24hh65gwlbi7wppcsa3q6m

Adaptive information retrieval: using a connectionist representation to retrieve and learn about documents

R. K. Belew
1989 SIGIR Forum  
The result is a representation of the consensual meaning of keywords and documents shared by some group of users.  ...  We argue that this associative representation is a natural generalization of traditional IR techniques, and that connectionist learning techniques are effective in this setting.  ...  I benefited a great deal from the brief time my path crossed his at the University of Michigan. I hope that I got at least part of his message right.  ... 
doi:10.1145/75335.75337 fatcat:qmdyu2lohjd27bn2tr3uvmmqou

Rule-based training of neural networks

Stan C. Kwasny, Kanaan A. Faisal
1991 Expert systems with applications  
This article describes a set of experiments with adaptive neural networks which explore two types of learning, deductive and inductive, in the context of a rule-based, deterministic parser of Natural Language  ...  We report on those experiences and draw some general conclusions that are relevant to knowledge engineering activities and maintenance of rule-based systems.  ...  Our work has shown some of the trade-offs between deductive and inductive learning.  ... 
doi:10.1016/0957-4174(91)90133-y fatcat:hvov2lzqyrdvtkk3d2bwvaqphi

An Analogy Anthology: Many Models, Some Data

Dorrit Billman
1996 Contemporary Psychology  
Thagard, of Induction: Processes of Inference, Learning, and Discovery and coeditor, with D. R. Shanks and D. L.  ...  First, many chapters trace out interesting and novel ways of combining connectionist and symbolic styles of computation or of ac- complishing cognitive tasks with a higher level favor using connectionist  ... 
doi:10.1037/004427 fatcat:b5jsfs7wyzen7k2toj6tsy2jwu
« Previous Showing results 1 — 15 out of 2,448 results