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 307 records · Page 17Linked to original sources
PubMed1997

Administration of DL-2-amino-5-phosphonovaleric acid (AP5) induces transient inhibition of reminder-activated memory retrieval in day-old chicks.

DL-2-Amino-5-phosphonovaleric acid (50 microM) administered immediately after a visual reminder presented to day-old chickens between 7.5 min and 24 h following a single trial passive avoidance learning task produced transient losses of memory on retention test, an effect not observed in the absence of a reminder or when the reminder was given 48 h post learning. The duration of the transient deficits decreased with increasing interval between training and the reminder trial. The time of onset of memory loss after the reminder trial appeared to increase with increasing interval between the training and the reminder trials. The results suggest that, for a period of at least up to 24 h after passive avoidance training, retrieval of memory may lead to processes which are sensitive to inhibition by the NMDA receptor antagonist AP5, with the duration of sensitivity post retrieval decreasing as the period of memory consolidation increases. The results extend previously reported findings and suggest the possibility that consolidation of a stable memorial representation of a learning experience may take over several days and may entail the concurrent laying down of a stable retrieval mechanism.

2-Amino-5-phosphonovalerate↗
PubMed2001

NMDA receptor antagonism blocks experience-dependent expansion of hippocampal "place fields".

In agreement with theories of sequence learning, hippocampal place representations expand asymmetrically during repeated route following. This behaviorally induced, experience-dependent expression of neuronal plasticity was blocked by the NMDA(R) antagonist CPP, suggesting that it may result from the temporal asymmetry and associative properties of LTP. NMDA(R) antagonism, however, had no effect on the range of the progressive shift of firing phase of hippocampal cells, relative to the theta rhythm, as the rat traverses the cell's "place field." Thus, when place fields normally expand with experience, the relationship between firing phase and position is altered, as predicted by models that account for "phase precession" on the basis of asymmetry of synaptic connection strengths. These effects of CPP mimic changes that occur during normal aging, suggesting mechanisms by which sequence learning deficits may arise in aged animals.

Animals↗
PubMed1998

A tennis serve and upswing learning robot based on bi-directional theory.

We experimented on task-level robot learning based on bi-directional theory. The via-point representation was used for 'learning by watching'. In our previous work, we had a robot learn kendama (a Japanese game) in order to demonstrate a single simple task. Our approach can be applied to a wide variety of motor behavior. However, some difficulties still remain. In this paper, we address two problems: (1) how to attain a final goal of complex movement when it consists of a sequence of subgoals, and (2) how to adapt to changes in behavior and the environment. To examine how to solve these problems, we propose two methods: (1) selecting the proper via-points for a control variable for each subgoal, and (2) re-estimating the relation between the via-points and the task during learning without conducting extra trials. We adopted a tennis serve and a pendulum upswing for our complicated tasks.

Journal Article↗
PubMed2001

Mediated conditioning and retrospective revaluation with LiCl then flavour pairings.

It is becoming accepted that the associative strength of a cue can change in its absence, despite this being difficult to explain using existing theories of Pavlovian conditioning. To investigate the influence of timing on learning about the representation of an absent cue, lithium chloride (LiCl) or a flavour previously paired with LiCl was presented in a distinctive context that had previously been paired with a neutral target flavour. The former treatment produced an aversion to the target flavour whether the LiCl was presented 10 min before, or immediately after, exposure to the context. However, presenting the flavour associate of LiCl created an aversion to the target flavour only if it had been presented 10 min after LiCl during initial training. This pattern of results cannot be explained in the terms of a simple timing account, and it is proposed that an explanation will require different associative rules operating in simultaneous and successive training schedules.

Animals↗
PubMed1998

Cortical plasticity: from synapses to maps.

It has been clear for almost two decades that cortical representations in adult animals are not fixed entities, but rather, are dynamic and are continuously modified by experience. The cortex can preferentially allocate area to represent the particular peripheral input sources that are proportionally most used. Alterations in cortical representations appear to underlie learning tasks dependent on the use of the behaviorally important peripheral inputs that they represent. The rules governing this cortical representational plasticity following manipulations of inputs, including learning, are increasingly well understood. In parallel with developments in the field of cortical map plasticity, studies of synaptic plasticity have characterized specific elementary forms of plasticity, including associative long-term potentiation and long-term depression of excitatory postsynaptic potentials. Investigators have made many important strides toward understanding the molecular underpinnings of these fundamental plasticity processes and toward defining the learning rules that govern their induction. The fields of cortical synaptic plasticity and cortical map plasticity have been implicitly linked by the hypothesis that synaptic plasticity underlies cortical map reorganization. Recent experimental and theoretical work has provided increasingly stronger support for this hypothesis. The goal of the current paper is to review the fields of both synaptic and cortical map plasticity with an emphasis on the work that attempts to unite both fields. A second objective is to highlight the gaps in our understanding of synaptic and cellular mechanisms underlying cortical representational plasticity.

Animals↗
PubMed2006

Operative time is a poor surrogate for the learning curve in laparoscopic colorectal surgery.

BACKGROUND: Previous studies have relied on conversion rate and operative time for construction of learning curves in laparoscopic colorectal surgery. The authors hypothesized that conversion rate and operative time were less important than complication and readmission rates in defining good outcomes and hence the learning curve. METHODS: A database of 287 consecutive laparoscopic colorectal resections from a single tertiary referral center was analyzed. Outcome measures included operative time, conversion rate, major and minor complications, length of hospital stay, and the 15- and 30-day hospital readmission rate. Data were analyzed both by surgeons and by quartile case numbers. RESULTS: A total of 151 right colectomies and 136 left colectomies were performed between 1995 and 2005. For both right and left colectomies, the conversion rate decreased in each of the first three quartiles, reaching a nadir of 0% for right colectomies and 3% for left colectomies in the third quartile. The conversion rates increased slightly in the fourth quartile. The operative time remained stable for three quartiles, then increased slightly in the fourth quartile. Two surgeons managed 199 of the 287 cases. Analysis of the two high-volume surgeons demonstrated that for left-sided resections, the surgeon with the shorter operative times had the higher major complication rate (13% vs 2%), overall complication rate (22% vs 2%), 30-day readmission rate (13% vs 0%), and length of stay (3.8 vs 3.1 days) (p < 0.05 for all comparisons). CONCLUSIONS: In this series, operative time failed to decrease with experience, and shorter operative times did not correlate with better clinical outcomes. The failure of operative time to decline with experience often reflects surgeons' willingness to attempt more difficult cases rather than an accurate representation of a "learning curve." Therefore, complication and readmission rates are more important than operative time and conversion rates for evaluating the learning curve and quality of laparoscopic colorectal surgery.

Adult↗
PubMed2006

Limitations on representation-mediated potentiation of flavour or odour aversions.

Odour aversion learning is often potentiated in the presence of flavour stimuli. Establishment of an aversion to an odour is greater when an odour + flavour compound is paired with illness than when the odour alone is paired with illness. Holland (1983) showed that under some circumstances auditory or olfactory stimuli previously paired with flavours may also potentiate odour aversion learning. The present experiments examined limitations on this representation-mediated potentiation of aversion learning. The results indicated that conditioned stimuli (CSs) that activate representations of potentiating cues are themselves immune to potentiation by other CS-activated representations, but remain susceptible to potentiation by their real stimulus associates.

Animals↗
PubMed2004

Cortical activity to vibrotactile stimulation: an fMRI study in blind and sighted individuals.

Blind individuals show visual cortex activity during Braille reading. We examined whether such cross-modal activations reflect processing somatosensory stimuli independent of language by identifying cortical activity during a one-back vibrotactile matching task. Three groups (sighted, early-onset, and late-onset [>12 years] blind) detected whether paired vibrations (25 and 100 Hz), delivered to the right index finger, differed in frequency. Three successive paired vibrations, followed by a no-stimulation interval, were presented in a long event-related design. A fixed effects average z-score analysis showed increased activity throughout the visuotopic visual cortex, where it was mostly restricted to foveal and parafoveal eccentricities. Early blind showed the most extensive distribution of activity. Late blind exhibited activity mostly in similar regions but with declining response magnitudes with age of blindness onset. Three sighted individuals had suprathreshold activity in V1 but negative responses elsewhere in visual cortex. Mixed effects ANOVA confirmed group distinctions in defined regions (V1, V3, V4v, V7, LOC, and MT). These results suggest cross-modal adaptation to tactile stimulation in visual cortex independent of language processes. All groups showed increased activity in left primary (S1) and bilateral second somatosensory areas, but without response magnitude differences between groups throughout sensorimotor cortex. Early blind showed the greatest spatial extent of S1 activity. Blind participants had more extensive bilateral activity in anterior intraparietal sulcus and supramarginal gyrus. Extensive usage of touch in Braille reading may underlie observed S1 expansions in the reading finger representation. In addition, learned attentiveness to touch may explain similar expansion of parietal tactile attention regions.

Adaptation, Physiological↗
PubMed1999

Effect of familiarity on the processing of human faces.

Most brain imaging studies on face perception have investigated the processing of unknown faces and addressed mainly the question of specific face processing in the human brain. The goal of this study was to highlight the effects of familiarity on the visual processing of faces. Using [15O]water 3D Positron Emission Tomography, regional cerebral blood flow distribution was measured in 11 human subjects performing an identical task (gender categorization) on both unknown and known faces. Subjects also performed two control tasks (a face recognition task and a visual pattern discrimination task). They were scanned after a training phase using videotapes during which they had been familiarized with and learned to recognize a set of faces. Two major results were obtained. On the one hand, we found bilateral activations of the fusiform gyri in the three face conditions, including the so-called fusiform-face area, a region in the right fusiform gyrus specifically devoted to face processing. This common activation suggests that different cognitive tasks performed on known and unknown faces require the involvement of this fusiform region. On the other hand, specific regional cerebral blood flow changes were related to the processing of known and unknown faces. The left amygdala, a structure involved in implicit learning of visual representations, was activated by the categorization task on unknown faces. The same task on known faces induced a relative decrease of activity in early visual areas. These differences between the two categorization tasks reveal that the human brain processes known and unknown faces differently.

Adult↗
PubMed1996

Modeling motor cortical operations by an attractor network of stochastic neurons.

Understanding the neural computations performed by the motor cortex requires biologically plausible models that account for cell discharge patterns revealed by neurophysiological recordings. In the present study the motor cortical activity underlying movement generation is modeled as the dynamic evolution of a large fully recurrent network of stochastic spiking neurons with noise superimposed on the synaptic transmission. We show that neural representations of the learned movement trajectories can be stored in the connectivity matrix in such a way that, when activated, a particular trajectory evolves in time as a dynamic attractor of the system while individual neurons fire irregularly with large variability in their interspike intervals. Moreover, the encoding of trajectories as attractors ensures high stability of the ensemble dynamics in the presence of synaptic noise. In agreement with neurophysiological findings, the suggested model can provide a wide repertoire of specific motor behaviors, whereas the number of specialized cells and specific connections may be negligibly small if compared with the whole population engaged in trajectory retrieving. To examine the applicability of the model we study quantitatively the relationship between local geometrical and kinematic characteristics of the trajectories generated by the network. The relationship obtained as a result of simulations is close to the '2/3 power law' established by psychophysical and neurophysiological studies.

Animals↗
PubMed1992

Functional organization of the auditory cortex: maps and mechanisms.

Recent studies have led to a better understanding of several aspects of the organization and physiological mechanisms involved in the processing of information in the auditory cortex. A wide range of approaches have revealed new information regarding the histochemistry, cortico-cortical connections, single-unit physiology, and functional spatial organization, as well as mechanisms and effects of representational and learning-induced cortical plasticity.

Animals↗
PubMed2005

Left thalamo-cortical network implicated in successful speech separation and identification.

The separation of concurrent sounds is paramount to human communication in everyday settings. The primary auditory cortex and the planum temporale are thought to be essential for both the separation of physical sound sources into perceptual objects and the comparison of those representations with previously learned acoustic events. To examine the role of these areas in speech separation, we measured brain activity using event-related functional Magnetic Resonance Imaging (fMRI) while participants were asked to identify two phonetically different vowels presented simultaneously. The processing of brief speech sounds (200 ms in duration) activated the thalamus and superior temporal gyrus bilaterally, left anterior temporal lobe, and left inferior temporal gyrus. A comparison of fMRI signals between trials in which participants successfully identified both vowels as opposed to when only one of the two vowels was recognized revealed enhanced activity in left thalamus, Heschl's gyrus, superior temporal gyrus, and the planum temporale. Because participants successfully identified at least one of the two vowels on each trial, the difference in fMRI signal indexes the extra computational work needed to segregate and identify successfully the other concurrently presented vowel. The results support the view that auditory cortex in or near Heschl's gyrus as well as in the planum temporale are involved in sound segregation and reveal a link between left thalamo-cortical activation and the successful separation and identification of simultaneous speech sounds.

Acoustic Stimulation↗
PubMed2004

Mind versus metabolism in the control of food intake and energy balance.

In a restrictive food environment, the homeostatic control system regulates body weight and adiposity with remarkable precision. However, this regulation appears to break down in many genetically predisposed individuals under conditions prevailing in the modern era characterized by a sedentary lifestyle and easy availability of large portions of palatable and calorically dense food. The nervous system is the main interface by which food-related environmental factors influence the regulatory process. Thus, focusing on the neural systems located in the telencephalon dealing with environmental factors, and on their connections with the homeostatic regulatory system distributed mainly in the hypothalamus and brainstem, should result in new drug targets and behavioral strategies for prevention and therapy. The structures providing this interface with the environment are involved in the initiation, procurement, and appetitive phases of ingestive behavior and associative learning before, during, and after the consummatory phase. It is thought that learned and unlearned representations of foods and food cues in the orbitofrontal and other cortical areas are filtered for affective/emotional value in the amygdala and for motivational salience in the nucleus accumbens/ventral striatum to initiate goal-directed motor programs. Internal state signals generated by the metabolic sensing mechanisms in the hypothalamus interact with each of these corticolimbic structures through reciprocal connections. While many projections from the hypothalamus contain the various "feeding peptides," the neurochemistry of projections to the hypothalamus has not been well characterized.

Animals↗
PubMed2004

Adaptive neural models of queuing and timing in fluent action.

In biological cognition, specialized representations and associated control processes solve the temporal problems inherent in skilled action. Recent data and neural circuit models highlight three distinct levels of temporal structure: sequence preparation, velocity scaling, and state-sensitive timing. Short sequences of actions are prepared collectively in prefrontal cortex, then queued for performance by a cyclic competitive process that operates on a parallel analog representation. Successful acts like ball-catching depend on coordinated scaling of effector velocities, and velocity scaling, mediated by the basal ganglia, may be coupled to perceived time-to-contact. Making acts accurate at high speeds requires state-sensitive and precisely timed activations of muscle forces in patterns that accelerate and decelerate the effectors. The cerebellum may provide a maximally efficient representational basis for learning to generate such timed activation patterns.

Adaptation, Physiological↗
PubMed2003

Neural timing of visual implicit categorization.

Most of the neuroimaging studies that have shown visual category-specific activations or categorization effects have been based on a subtractive approach. In the present study, we investigated, by means of EEG, not only the net result of the categorization but also the dynamics of the process. Subjects had to perform a target detection task throughout an image set of distractors belonging to six categories: letters, geometrical figures, faces, tools, structured textures and Asiatic characters. Multivariate analyses were performed on the responses to the non-target stimuli according to their category. Categorical neural responses were only obtained on P2 latencies and N2 amplitudes. This result suggests that there are no differences in the first stage of the implicit categorization of the distractors (visual analysis and proximal stimulus representation elaboration from 100 to 220 ms) and that differences appear between 220 and 280 ms (matching to structural representations). Over-learned stimuli (e.g. letters) elicited the shortest P2 latency, contrasting with unknown categories (e.g. Asiatic characters) that revealed the longest P2 latencies and flattened N2 waves. Categorical differences indicate that the more a subject knows about an object, the less cognitive resources are used. In conclusion, our results suggest that a reduction in neural activity could reflect an improved accuracy in cognitive and cortical processing.

Adult↗
PubMed2006

Developmental dyspraxia is not limited to imitation in children with autism spectrum disorders.

Impaired imitation of skilled gestures is commonly reported in autism. Questions, however, remain as to whether impaired imitation is associated with a more generalized deficit in performance of gestures consistent with a dyspraxia and whether the pattern of errors differs from that observed in typically developing children. To address these questions, praxis in 21 high-functioning children with autism spectrum disorders (ASD) was compared with 24 typically developing controls using a traditional approach in which performance was evaluated through detailed examination of error types. Children with ASD produced significantly fewer correct responses not only during Gesture to Imitation, but also during Gesture to Command and with Tool Use. The pattern of errors in ASD was similar to that of controls with spatial errors being most common in both groups; however, body-part-for-tool errors were more common in children with ASD, suggesting dyspraxia is not entirely attributable to motor deficits. The findings suggest that autism is associated with a generalized praxis deficit, rather than a deficit specific to imitation. In a developmental disorder such as autism, the findings may reflect abnormalities in frontal/parietal-subcortical circuits important for acquisition (i.e., learning) of sensory representations of movement and/or the motor sequence programs necessary to execute them.

Apraxias↗
PubMed2026

CoxFormer enables spatial omics inference with multimodal generative modeling.

Gene co-expression maps transcriptome-wide gene-gene relationships, yet high-quality estimates cover less than half the genome. Meanwhile, spatial omics either profiles restricted in situ panels or lacks cellular resolution. Extending co-expression transcriptome-wide could overcome these limitations by inferring unassayed gene expression at subcellular resolution. Here we show that CoxFormer integrates literature-derived gene knowledge with co-expression networks from bulk tissues and large-scale single-cell atlases to learn 512-dimensional representations for 32,016 human genes. These embeddings capture functional gene relationships and serve as a generative prior for spatial inference across platforms and modalities. Without requiring a matched single-cell RNA-sequencing reference, CoxFormer supports four applications beyond measured genes: histology-based expression imputation, gene activity prediction from chromatin accessibility, subcellular super-resolution inference, and pathological region detection. Together, CoxFormer extends gene embedding from gene- and cell-level tasks to whole-transcriptome spatial inference, providing a unified framework for biological analysis beyond the limited gene coverage of current spatial omics technologies.

Humans↗
PubMed1992

IPSA-Inductive Protein Structure Analysis.

The Inductive Structure Protein Analysis (IPSA) project presents a new method for investigating protein structure. IPSA includes the creation of a new database which was designed specifically for the analysis of protein structure by statistics and machine learning. The Protein Representation Language (PRL) database includes explicit and symbolic representations of geometrical, topological and chemophysical information about secondary structures and the relationships between secondary structures. The IPSA methodology consists of: the use of PRL information to produce a new database of examples of secondary structures which associate together (examples of possible super-secondary structures); then the use of a variety of clustering techniques to produce a consensus clustering of these examples (super-secondary structures); these super-secondary structures are finally examined to uncover any biological features of significance. We have applied this method to find simple super-secondary structures consisting of pairs of alpha-helices. We found four well-defined super-secondary structures, one formed exclusively by long range interactions, and another in association with an additional element of secondary structure (alpha t alpha-motif). Examinations were carried out using homologous pairs and conformational fits which confirm our clustering.

Cluster Analysis↗