What connectionist models learn: Learning and representation in connectionist networks

1990 ◽  
Vol 13 (3) ◽  
pp. 471-489 ◽  
Author(s):  
Stephen José Hanson ◽  
David J. Burr
Keyword(s):  
Computational Approach ◽  
Complex Interaction ◽  
Connectionist Models ◽  
Promising Alternative ◽  
Rule Based ◽  
Symbol Processing ◽  
Connectionist Networks ◽  
Connectionist Approach ◽  
The Brain ◽  
Explicit Rule

AbstractConnectionist models provide a promising alternative to the traditional computational approach that has for several decades dominated cognitive science and artificial intelligence, although the nature of connectionist models and their relation to symbol processing remains controversial. Connectionist models can be characterized by three general computational features: distinct layers of interconnected units, recursive rules for updating the strengths of the connections during learning, and “simple” homogeneous computing elements. Using just these three features one can construct surprisingly elegant and powerful models of memory, perception, motor control, categorization, and reasoning. What makes the connectionist approach unique is not its variety of representational possibilities (including “distributed representations”) or its departure from explicit rule-based models, or even its preoccupation with the brain metaphor. Rather, it is that connectionist models can be used to explore systematically the complex interaction between learning and representation, as we try to demonstrate through the analysis of several large networks.

scholarly journals Veto and Vacillation: A Neural Precursor of the Decision to Withhold Action

2014 ◽  
Vol 26 (2) ◽  
pp. 296-304 ◽  
Author(s):  
Erman Misirlisoy ◽  
Patrick Haggard
Keyword(s):  
Human Behavior ◽  
Temporal Pattern ◽  
Neural Precursor ◽  
Rule Based ◽  
Action Sequences ◽  
Readiness Potentials ◽  
Preparatory Activity ◽  
The Brain ◽  
Motor Areas

The capacity to inhibit a planned action gives human behavior its characteristic flexibility. How this mechanism operates and what factors influence a decision to act or not act remain relatively unexplored. We used EEG readiness potentials (RPs) to examine preparatory activity before each action of an ongoing sequence, in which one action was occasionally omitted. We compared RPs between sequences in which omissions were instructed by a rule (e.g., “omit every fourth action”) and sequences in which the participant themselves freely decided which action to omit. RP amplitude was reduced for actions that immediately preceded a voluntary omission but not a rule-based omission. We also used the regular temporal pattern of the action sequences to explore brain processes linked to omitting an action by time-locking EEG averages to the inferred time when an action would have occurred had it not been omitted. When omissions were instructed by a rule, there was a negative-going trend in the EEG, recalling the rising ramp of an RP. No such component was found for voluntary omissions. The results are consistent with a model in which spontaneously fluctuating activity in motor areas of the brain could bias “free” decisions to act or not.

scholarly journals New Technologies of Motor Control and Rehabilitation

2020 ◽  
Vol 1 (1) ◽  
Author(s):  
Albertas Skurvydas
Keyword(s):  
Motor Control ◽  
New Technologies ◽  
Modern Science ◽  
Computational Approach ◽  
Neural Cells ◽  
Motor Rehabilitation ◽  
Principal Mechanism ◽  
Brain Stroke ◽  
Computational Paradigm ◽  
The Brain

Modern paradigms of motor control and rehabilitation are analyzed in the paper. Two main paradigms, i. e. computational approach and dynamical system approach are engaged in rivalry in motor control and learning research at present. From the standpoint of computational paradigm the principal mechanism of motor control and learning consists in the ability of the brain “to calculate” (acting as some kind of biological computer). According to the paradigm of dynamical systems the mechanism of motor control is time dependent. In other words, it can be different each time. The main principles of motor control and properties of movements are given considerable attention in the paper. Besides, modern methods of motor rehabilitation after stroke are emphasized in the paper. Fitting of neuroprosthesis and restoration of damaged neural cells are significant maiden steps in modern science. The scientists are engaged in search for: a) constraining such mechanism prosthesis that would submit to the efforts of human will and b) restoring neural cells damaged because of the brain stroke suffered.Keywords: motor control, rehabilitation, stroke.

Re-assembling the brain: Are cell assemblies the brain's language for recovery of function?

1999 ◽  
Vol 22 (2) ◽  
pp. 284-284 ◽  
Author(s):  
Chris Code
Keyword(s):  
Cognitive Functions ◽  
Recovery Of Function ◽  
Cell Assembly ◽  
Connectionist Models ◽  
Cell Assemblies ◽  
The Brain

Holistically ignited Hebbian models are fundamentally different from the serially organized connectionist implementations of language. This may be important for the recovery of language after injury, because connectionist models have provided useful insights into recovery of some cognitive functions. I ask whether cell assembly modelling can make an important contribution and whether the apparent incompatibility with successful connectionist modelling is a problem.

Extracting higher-level relationships in connectionist models

1997 ◽  
Vol 20 (1) ◽  
pp. 77-77
Author(s):  
Gary F. Marcus
Keyword(s):  
Recurrent Network ◽  
Higher Order ◽  
Connectionist Models ◽  
Connectionist Networks ◽  
Statistical Regularities

Connectionist networks excel at extracting statistical regularities but have trouble extracting higher-order relationships. Clark & Thornton suggest that a solution to this problem might come from Elman (1993), but I argue that the success of Elman's single recurrent network is illusory, and show that it cannot in fact represent abstract relationships that can be generalized to novel instances, undermining Clark & Thornton's key arguments.

scholarly journals The Skin Microbiota and Itch: Is There a Link?

2020 ◽  
Vol 9 (4) ◽  
pp. 1190
Author(s):  
Hei Sung Kim ◽  
Gil Yosipovitch
Keyword(s):  
Neuronal Activity ◽  
Inflammatory Cells ◽  
Complex Interaction ◽  
Nerve Endings ◽  
Skin Microbiota ◽  
Pain Research ◽  
Unpleasant Sensation ◽  
Major Player ◽  
The Brain ◽  
Chemical Mediators

Itch is an unpleasant sensation that emanates primarily from the skin. The chemical mediators that drive neuronal activity originate from a complex interaction between keratinocytes, inflammatory cells, nerve endings and the skin microbiota, relaying itch signals to the brain. Stress also exacerbates itch via the skin–brain axis. Recently, the microbiota has surfaced as a major player to regulate this axis, notably during stress settings aroused by actual or perceived homeostatic challenge. The routes of communication between the microbiota and brain are slowly being unraveled and involve neurochemicals (i.e., acetylcholine, histamine, catecholamines, corticotropin) that originate from the microbiota itself. By focusing on itch biology and by referring to the more established field of pain research, this review examines the possible means by which the skin microbiota contributes to itch.

scholarly journals The color phi phenomenon: Not so special, after all?

2021 ◽  
Vol 17 (9) ◽  
pp. e1009344
Author(s):  
Lars Keuninckx ◽  
Axel Cleeremans
Keyword(s):  
Neural Network ◽  
Recurrent Neural Network ◽  
Sensory Processing ◽  
Time Reversal ◽  
Echo State Network ◽  
Connectionist Models ◽  
Long Distance ◽  
Minimal Set ◽  
The Brain

We show how anomalous time reversal of stimuli and their associated responses can exist in very small connectionist models. These networks are built from dynamical toy model neurons which adhere to a minimal set of biologically plausible properties. The appearance of a “ghost” response, temporally and spatially located in between responses caused by actual stimuli, as in the phi phenomenon, is demonstrated in a similar small network, where it is caused by priming and long-distance feedforward paths. We then demonstrate that the color phi phenomenon can be present in an echo state network, a recurrent neural network, without explicitly training for the presence of the effect, such that it emerges as an artifact of the dynamical processing. Our results suggest that the color phi phenomenon might simply be a feature of the inherent dynamical and nonlinear sensory processing in the brain and in and of itself is not related to consciousness.

scholarly journals Morpho-Phonetic Effects in Speech Production: Modeling the Acoustic Duration of English Derived Words With Linear Discriminative Learning

2021 ◽  
Vol 12 ◽  
Author(s):  
Simon David Stein ◽  
Ingo Plag
Keyword(s):  
Speech Production ◽  
Morphological Structure ◽  
Computational Approach ◽  
Natural Speech ◽  
Discriminative Learning ◽  
Promising Alternative ◽  
Learning Network ◽  
Speech Data ◽  
Production Modeling

Recent evidence for the influence of morphological structure on the phonetic output goes unexplained by established models of speech production and by theories of the morphology-phonology interaction. Linear discriminative learning (LDL) is a recent computational approach in which such effects can be expected. We predict the acoustic duration of 4,530 English derivative tokens with the morphological functions DIS, NESS, LESS, ATION, and IZE in natural speech data by using predictors derived from a linear discriminative learning network. We find that the network is accurate in learning speech production and comprehension, and that the measures derived from it are successful in predicting duration. For example, words are lengthened when the semantic support of the word's predicted articulatory path is stronger. Importantly, differences between morphological categories emerge naturally from the network, even when no morphological information is provided. The results imply that morphological effects on duration can be explained without postulating theoretical units like the morpheme, and they provide further evidence that LDL is a promising alternative for modeling speech production.

scholarly journals A Survey in Implementation and Applications of Electroencephalograph (EEG)-Based Brain-Computer Interface

2021 ◽  
Vol 39 (7) ◽  
pp. 1117-1132
Author(s):  
Samaa S. Abdulwahab ◽  
Hussain K. Khleaf ◽  
Manal H. Jassim
Keyword(s):  
Brain Computer Interface ◽  
Physical World ◽  
The Body ◽  
Computer Interface ◽  
Signal Acquisition ◽  
Rule Based ◽  
The Past ◽  
Brain Signals ◽  
Translation Systems ◽  
The Brain

A Brain-Computer Interface (BCI) is an external system that controls activities and processes in the physical world based on brain signals. In Passive BCI, artificial signals are automatically generated by a computer program without any input from nerves in the body. This is useful for individuals with mobility issues. Traditional BCI has been dependent only on recording brain signals with Electroencephalograph (EEG) and has used a rule-based translation algorithm to generate control commands. These systems have developed very accurate translation systems. This paper is about the different methods for adapting the signals from the brain. It has been mentioned that various kinds of surveys in the past to serve the purpose of the present research. This paper shows a simple and easy analysis of each technique and its respective benefits and drawbacks, including signal acquisition, signal pre-processing, feature classification and classification. Finally,  discussed is the application of EEG-based BCI.

Sign in / Sign up

Export Citation Format

Share Document