Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “representation learning”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 505 records · Page 28Linked to original sources
PubMed1993

Taxonomy of subjective phenomena: a neuropsychological basis of functional assessment of ischemic or traumatic brain lesions.

A proper evaluation of functional competence after central lesions has to be based on a classification of functions that one can agree upon. It is a sad fact in neuropsychology that such a classification is not available. An attempt will be made to discuss such a classification (or taxonomy) that might be useful. The basic idea is that elementary psychological functions are evolutionary products whose availability is dependent on the functional integrity of neuronal modules. Such modules are embedded neuronal mechanisms that are linked to localized structures or distributed neuronal algorithms. Constancy of interindividual loss of psychological functions associated with lesions of modules can be used to define a catalogue of functions. Using this principle one can differentiate four areas of psychological functions that are represented in a modular fashion. These areas are stimulus representations ("perception"), processing of information ("learning and memory"), evaluation of information (for instance by emotions), and finally action or reaction. Functional competence is, however, not only described by the potential availability of elementary psychological functions, but also by formal aspects, i.e. how functions are made available. Such formal aspects refer to activation and in particular to temporal problems of neuronal processing. A particular "time machine" will be discussed which is essential for functional competence. Central lesions may either effect the what of functions or the how of functions. A differentiation between these material and formal aspects of functional competence are essential with respect to recovery or restitution of function.

Algorithms↗
PubMed1997

Unilateral striatal lesions in the cat disrupt well-learned motor plans in a GO/NO-GO reaching task.

We examined the changes in learned and spontaneous motor behavior after a unilateral excitotoxin lesion of the neostriatum. Cats were trained to perform a sensory-cued GO/NO-GO reaching task. Success rate, reaction time, movement speed and kinematic patterns were used to characterize motor system properties. In addition, motor properties before and after the lesion were compared by clinical neurological examinations and video tape observations of free-range behavior. We found that in normal animals motor performance in the task was fluent, highly automatic and skillful with consistent patterns from trial to trial and day to day. The striatal lesion resulted in a marked impairment in the animals' ability to perform the automatic response to the sensory cues in the motor task. In contrast, sensorimotor behavior in contexts apart from the task was altered minimally, with changes that were often difficult to detect. The animals recovered their ability to perform the task gradually, although they never reached prelesion performance levels in up to 24 weeks of evaluation. The animals had difficulty making reaching movements in GO trials and, in NO-GO trials failures to withhold movements were more frequent. Failures were due to a specific inability to execute previously well-learned movements in response to cues and not to an inability to recognize and interpret the cues. The lesion effects were restricted to the automatic motor response to the learned cues, as the animals could make reaching movements to the target without obvious impairment in response to novel stimuli. They also made similar spontaneous movements apart from the motor task that appeared to be unimpaired. The unique motor style and strategies that characterized the behavior of individual animals prior to the lesion were still evident after the lesion, even though they were superimposed on lower success rates and slower movement speeds. Our findings suggest that the basal ganglia facilitate the fluent and rapid execution of sequences of well-learned sensorimotor behavior, but the representations of motor plans are not stored in the basal ganglia.

Animals↗
PubMed1996

Gaussian ARTMAP: A Neural Network for Fast Incremental Learning of Noisy Multidimensional Maps.

A new neural network architecture for incremental supervised learning of analog multidimensional maps is introduced. The architecture, called Gaussian ARTMAP, is a synthesis of a Gaussian classifier and an adaptive resonance theory (ART) neural network, achieved by defining the ART choice function as the discriminant function of a Gaussian classifier with separable distributions, and the ART match function as the same, but with the distributions normalized to a unit height. While Gaussian ARTMAP retains the attractive parallel computing and fast learning properties of fuzzy ARTMAP, it learns a more efficient internal representation of a mapping while being more resistant to noise than fuzzy ARTMAP on a number of benchmark databases. SSeveral simulations are presented which demonstrate that Gaussian ARTMAP consistently obtains a better trade-off of classification rate to number of categories than fuzzy ARTMAP. Results on a vowel classification problem are also presented which demonstrate that Gaussian ARTMAP outperforms many other classifiers. Copyright 1996 Elsevier Science Ltd

Journal Article↗
PubMedDate not supplied

The neural representation of language in users of American Sign Language.

UNLABELLED: Studies of American Sign Language (ASL) offer unique insights into the fundamental properties of human language. Neurolinguistic studies explore the effects of left and right hemisphere lesions on the production and comprehension of signed language. Following damage to the left hemisphere perisylvian regions, signers, like users of spoken languages, exhibit frank aphasic disturbances. Sign language paraphasia illustrates the linguistic specificity of impairment. A case study involving cortical stimulation mapping (CSM) in a deaf signer provides evidence for the specialization of Broca's area in sign language production. The effects of right hemisphere damage highlight the specialized properties of sign language use. Data from functional magnetic resonance imaging (fMRI) of deaf signers confirm the importance of left hemisphere language structures in the use of signed language, but also reveal the contributions of right hemisphere regions to the processing of ASL. These studies provide new insights into the complementary roles of biology and environment in language representation in the human brain. LEARNING OUTCOMES: As a result of this activity, the participant will read studies of aphasia in users of signed language and a discussion of neurolinguistic studies of paraphasia in ASL. The participant will examine the role of the right hemisphere in language use and findings from a functional imaging study of sentence processing in ASL and English.

Aphasia, Broca↗
PubMed2005

Is the cerebellum ready for navigation?

Spatial navigation required the acquisition of at least two complementary processes: the organization of the spatial representation of the environment (declarative learning) and the acquisition of a motor behaviour adapted to the specific context (procedural learning). The potential role of the cerebellum in spatial navigation is part of the debate concerning its role in cognitive function. Experiments ranging from cerebellar patients to animal models have indicated that cerebellar damage affects the processing of spatial information. The main unresolved issue concern the interpretation of these deficits. Is the cerebellum involved in both declarative and procedural components of navigation? Could all deficits in navigation paradigms be interpreted by a deficit in a motor-dependant process? The purpose of this review is to examine different results coming from anatomical data, experimental paradigms and models in order to give a critical answer to this question.

Animals↗
PubMed2003

Auditory and visual automatic attention deficits in developmental dyslexia.

Several studies have provided evidence for a phonological deficit in developmental dyslexia. However, recent studies provide evidence for a multimodal temporal processing deficit in dyslexia. In fact, dyslexics show both auditory and visual abnormalities, which could result from a more general problem in the perceptual selection of stimuli. Here we report the results of a behavioral study showing that children with dyslexia have both auditory and visual deficits in the automatic orienting of spatial attention. These findings suggest that a deficit of selective spatial attention may distort the development of phonological and orthographic representations that is essential for learning to read.

Acoustic Stimulation↗
PubMed1999

Connectionist modeling of speech perception.

Connectionist models of perception and cognition, including the process of deducing meaningful messages from patterns of acoustic waves emitted by vocal tracts, are developed and refined as human understanding of brain function, psychological processes, and the properties of massively parallel architectures advances. The present article presents several important contributions from diverse points of view in the area of connectionist modeling of speech perception and discusses their relative merits with respect to specific theoretical issues and empirical findings. TRACE, the Elman/Norris net, and Adaptive Resonance Theory constitute pivotal points exemplifying overall modeling success, progress in temporal representation, and plausible modeling of learning, respectively. Other modeling efforts are presented for the specific insights they offer, and the article concludes with a discussion of computational versus dynamic modeling of phonological processes.

Brain↗
PubMed2006

Sociocultural contexts for the early development of semiotic production.

Children constantly encounter signs during cultural practices, although many theories do not fully acknowledge sociocultural aspects of semiotic development. The author examines research on cultural practices and contexts in which children learn to produce signs involving representational drawing and pretend play. This work is contrasted with more individualistic views of semiotic development that fail to adequately address sociocultural aspects of semiotic functioning to varying degrees. The author also presents a theoretical model for understanding the structure of any sign system and for comparing semiotic systems, using examples from the drawing and pretense literatures. It is proposed that the sign-making practices in which children participate are shaped by a complex hierarchy of conventions.

Art↗
PubMed2006

Lack of set size effects in spatial updating: Evidence for offline updating.

Four experiments required participants to keep track of the locations of (i.e., update) 1, 2, 3, 4, 6, 8, 10, or 15 target objects after rotating. Across all conditions, updating was unaffected by set size. Although some traditional set size effects (i.e., a linear increase of latency with memory load) were observed under some conditions, these effects were independent of the updating process. Patterns of data and participant strategies were inconsistent with the common view of spatial updating as an online process. Instead, the authors concluded that participants formed enduring, long-term memory representations of the layouts at learning that were used to reconstruct spatial information about the layouts as needed (i.e., offline updating). These results support M. Amorim, S. Glasauer, K. Corpinot, and A. Berthoz's (1997) 2-system model of spatial updating that includes both online and offline updating.

Adolescent↗
PubMed2006

Neural mechanisms of birdsong memory.

The process through which young male songbirds learn the characteristics of the songs of an adult male of their own species has strong similarities with speech acquisition in human infants. Both involve two phases: a period of auditory memorization followed by a period during which the individual develops its own vocalizations. The avian 'song system', a network of brain nuclei, is the probable neural substrate for the second phase of sensorimotor learning. By contrast, the neural representation of song memory acquired in the first phase is localized outside the song system, in different regions of the avian equivalent of the human auditory association cortex.

Animals↗
PubMed2025

Integration of single cell multiomics data by deep transfer hypergraph neural network.

Multi-omics characterization of individual cells offers remarkable potential for analyzing the dynamics and relationships of gene regulatory states across millions of cells. How to integrate multimodal data is an open problem, existing integration methods struggle with accuracy and modality-specific biological variation retention. In this paper, we present scHyper (scalable, interpretable machine learning for single cell integration), a low-code and data-efficient deep transfer model designed for integrating paired and unpaired single-cell multimodal data. We benchmark scHyper against datasets from different multimodal data. ScHyper learns a low-dimensional representation and aligns the covariance matrices of the measured modalities, achieving high accuracy even with large scale atlas-level datasets with low memory and computational time across different cell lines, shedding light on regulatory relationships between different types of omics. Altogether, we show that scHyper is a versatile and robust tool for cell-type label transfer and integration from multimodal single-cell datasets.

Single-Cell Analysis↗
PubMed2001

Transcranial magnetic stimulation as a tool for cognitive studies.

Transcranial Magnetic Stimulation (TMS) is a tool for the non-invasive stimulation of the human brain. It allows the activation of arbitrary sites of the superficial cortex and, combined with other brain-imaging techniques such as EEG, PET, and fMRI, it can be used to evaluate cortical excitability and connectivity. This is of major importance in, for example, the study of cognitive processes such as language, learning, memory and self-representation, which are thought to be represented in multiple brain areas. The mechanisms of action of TMS are known on a basic level, but its effect on the activation state of brain tissue is still poorly understood. Clinical applications of TMS have also been proposed and guidelines for its safe use drafted.

Brain↗
PubMed2005

Adaptation to visuomotor rotations remaps movement vectors, not final positions.

When exposed to novel visuomotor rotations, subjects readily adapt reaching movements, such that the virtual display of the hand is brought to the target. Whereas this clearly reflects remapping of the relationship between hand movements and the visual display, the nature of this remapping is not well understood. We now examine whether such adaptation results in remapping of the position of the visually displayed target and the final limb position or between the target vector and the movement vector. The latter is defined relative to a starting position, whereas the former should be independent of the starting position. Subjects first adapted to a 30 degrees rotation during reaching movements made from a single starting location to four different target locations. After adaptation, generalization trials were introduced, during which reaching movements were made under the same visual rotation condition but started from one of two locations outside the practiced workspace. These trials were directed to either the previously practiced targets or new targets that reflected the direction and distance of the practiced trials. Generalization was greatest for movements made in similar directions, regardless of changes in spatial location. Most significantly, when reaching to the previously adapted targets, subjects did not reach to the previously learned limb positions but rather to positions that reflected a near 30 degrees rotation of the new target vector. These results indicate that learned visuomotor rotations remap the representations of movement vectors and not final positions of the limb in the workspace.

Acclimatization↗
PubMedDate not supplied

[Study of the relationships between self-injurious behavior and pain reactivity in infantile autism].

Autism can be considered as an early general developmental disorder, characterized by problems of social interaction, problems of verbal and non verbal communication, and behavioral or ideational stereotypes. However, within autism we observe a clinical heterogeneity of autistic disorders which suggests the possibility of autistic subtypes. Several authors hypothesize an analgesia among autistic children; this analgesia may be related to self-mutilation found among autistics. The current research had two objectives: 1) to develop and validate evaluation tools for measuring aggression directed towards the self (Yale-Paris Self-Injurious Behavior Scale: YAPA SIB) and pain reactivity (Pre-Linguistic Behavioral Pain Reactivity Scale: PLBPRS); instruments appropriate for autistics and capable of showing different behavioral sub-types; 2) to study in 80 autistic children pain reactivity, self-injurious behavior, and their relation in different observational situations. The results show that the scales of self-injurious behavior and pain reactivity have good discriminative capacity, good test-retest reliability, and good validity. The results suggest additionally that the apparent decreased pain reactivity observed in autistics does not derive from a real analgesia but from a different mode of pain expression related to difficulties with verbal communication, body representation and certain cognitive disorders (learning disorders, problems representing sensations and emotions, problems establishing cause-effect relationships). Additionally, there is a significant relationship between certain self-injurious behaviors and the apparent reduced pain reactivity. Interpretations of this result are presented and the possible role of stress in autism is discussed.

Autistic Disorder↗
PubMed2005

A nonlinear adaptive fuzzy approximator technique with its application to prediction of non-stationary EEG dynamics and estimation of single-sweep evoked potentials.

In this paper, we apply a fast training paradigm to the optimization of fuzzy approximator and a nonlinear adaptive fuzzy approximator (NAFA) is constructed. Using TSK fuzzy rules, the structured knowledge along with numerical information are parameterized and utilized in the NAFA, which can be easily configured as a multi-layer network when its transparency is desired. We propose a fast training paradigm, which is actually a combination of Kalman filtering and LMS adaptation, to optimize the linear and nonlinear parameters of the NAFA separately. The NAFA is characterized by concise representation of structured knowledge, fast learning capability, as well as universal approximation property. The NAFA is applied to forecast the non-stationary EEG time-series and to estimate single-sweep evoked potentials (EPs). The corresponding simulation results are given. It is concluded that the NAFA technique can provide efficient nonlinear separation of single-sweep EPs, which allows for quantitative examination of the cross-trial variability of clinical EPs.

Artifacts↗
PubMed2005

Behavioral evidence that segregation and representation are dissociable hippocampal functions.

Hippocampal activity is thought to encode spatial representations in a distributed associative network. This idea predicts that partial hippocampal lesions would spare acquisition and impair retrieval of a place response as long as enough connections remained intact to encode associations. Water maze experiments supported the predictions, but the prediction of impaired retrieval was not supported when tetrodotoxin (TTX) was injected into one hippocampus and rats were tested in a place avoidance task on a rotating arena with shallow water. The rotation dissociated relevant distal stimuli from irrelevant self-motion stimuli. To explain the discrepancy, we hypothesized that the segregation of relevant and irrelevant stimuli and stimuli association into representations are distinct hippocampus-dependent operations, and whereas associative representation is more sensitive to disruption during retrieval than learning, stimulus segregation is more sensitive to disruption during learning than during retrieval. The following predictions were tested: (1) the TTX injection would spare learning but (2) impair retrieval of a place response in the water maze, which has a high associative representational demand but a low demand for segregation; (3) the injection would impair learning but (4) spare retrieval of place avoidance in the rotating arena filled with water, which has a high demand for stimulus segregation but a low associative representational demand. All four predictions were confirmed. The hypothesis also explains the pattern of sparing and impairment after the TTX injection in other place avoidance task variants, leading us to conclude that stimulus separation and association representation are dissociable functions of the hippocampus.

Analysis of Variance↗
PubMed1989

Learning arm kinematics and dynamics.

In this review I have discussed how the form of representation used in internal models of the motor apparatus affects how and what a system can learn. Tabular models and structured models have benefits and drawbacks. Structured models incorporate knowledge of the structure of the controlled motor apparatus. If that knowledge is correct, or close to the actual system structure, the structured models will support global generalization and rapid, efficient learning. Tabular models can play an important role in learning to control systems when either the system structure is not known or only known approximately. Tabular models are general and flexible. Techniques for combining these different representations to attain the benefits of both are currently under investigation. In the control of multijoint systems such as the human arm, internal models of the motor apparatus are necessary to interpret performance errors. In the study of movements restricted to one joint, the problem of interpreting performance errors is greatly simplified and often overlooked, as performance errors can usually be related to command corrections by a single gain. When multijoint movements of the same motor systems are examined, however, the complex nature of the control and coordination problems faced by the nervous system become evident, as well as the sophistication of the brain's solutions to these problems. Recent progress in the understanding of adaptive control of eye movements provides a good example of this (Berthoz & Melvill-Jones 1985). Experimental studies of the psychophysics of motor learning can play an important role in bridging the gap between computational theories of how abstract motor systems might learn and physiological exploration of how actual nervous systems implement learning. Quantitative analyses of the patterns of motor learning of biological systems may help distinguish alternative hypotheses about the representations used for motor control and learning. What a system can and cannot learn, the amount of generalization, and the rate of learning give clues as to the underlying performance architecture. It is also important to know the actual performance level of the motor system (Loeb 1983). Different proposed control strategies will be able to attain different performance levels, and the use of simplifying control strategies may be evident in the control and learning performance of motor systems.

Arm↗
PubMed2006

Transcranial magnetic stimulation and the motor learning-associated cortical plasticity.

It has been well established that repetitive motor performance and skill learning alter the functional organization of human corticomotoneuronal system. Over the past decade, transcranial magnetic stimulation (TMS) has helped to demonstrate motor practice and learning-related changes in corticomotoneuronal excitability and representational plasticity. It has also provided some insights into the mechanisms underlying such plasticity. TMS-derived indices show that motor practice, skill acquisition and learning are associated with an increase in cortical excitability and a modulation of intracortical inhibition partly related to the amount of GABA-related inhibition. It has been suggested that these changes in excitability might be related to learning and motor memory formation in the motor cortex. However, it has proved difficult to relate different aspects of TMS-derived representational plasticity with specific behavioral outcomes. A better understanding of the relationship between TMS measurements of practice-related cortical plasticity and underlying mechanisms, in the context of associated changes in behavior, will facilitate the development of techniques and protocols that will allow predictable modulation of cortical plasticity in health and disease.

Animals↗