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 1,135 records · Page 63Linked to original sources
PubMed2002

Force adaptation transfers to untrained workspace regions in children: evidence for developing inverse dynamic motor models.

When humans perform goal-directed arm movements under the influence of an external damping force, they learn to adapt to these external dynamics. After removal of the external force field, they reveal kinematic aftereffects that are indicative of a neural controller that still compensates the no longer existing force. Such behavior suggests that the adult human nervous system uses a neural representation of inverse arm dynamics to control upper-extremity motion. Central to the notion of an inverse dynamic model (IDM) is that learning generalizes. Consequently, aftereffects should be observable even in untrained workspace regions. Adults have shown such behavior, but the ontogenetic development of this process remains unclear. This study examines the adaptive behavior of children and investigates whether learning a force field in one hemifield of the right arm workspace has an effect on force adaptation in the other hemifield. Thirty children (aged 6-10 years) and ten adults performed 30 degrees elbow flexion movements under two conditions of external damping (negative and null). We found that learning to compensate an external damping force transferred to the opposite hemifield, which indicates that a model of the limb dynamics rather than an association of visited space and experienced force was acquired. Aftereffects were more pronounced in the younger children and readaptation to a null-force condition was prolonged. This finding is consistent with the view that IDMs in children are imprecise neural representations of the actual arm dynamics. It indicates that the acquisition of IDMs is a developmental achievement and that the human motor system is inherently flexible enough to adapt to any novel force within the limits of the organism's biomechanics.

Adaptation, Physiological↗
PubMed1991

Local adaptations of two naturally occurring neuronal conductances, gK + (A) and gK + (Ca), allow for associative conditioning and contiguity judgements in artificial neural networks.

Features of two potassium conductances implicated in the acquisition of conditioned reflexes, the slow calcium dependent conductance (gK + (Ca] and the fast transient conductance (gK + (A], were incorporated into a 6 x 6 element artificial neural network. Adaptive algorithms derived from observations of cortical neurons during associative learning changed gK + (A) in proportion to the product of this current and an EPSP-induced second messenger concentration, and changed gK + (Ca) as a function of a spike-induced second messenger concentration. This network concurrently acquired two distinct representations in response to presentation of stimuli: one resembled associative conditioning (defined in terms of its sensitivity to forward pairing vs. simultaneous or backward pairing); the other reflected contiguous pairings of stimuli. The acquisition of one representation did not markedly interfere with acquisition of the other. This network may accordingly serve as an example of a self-organizing system which minimizes the postulated inherent cross talk between functionally dissiminar representations (Minsky and Papert 1988).

Acclimatization↗
PubMed2006

Enhancing text categorization with semantic-enriched representation and training data augmentation.

OBJECTIVE: Acquiring and representing biomedical knowledge is an increasingly important component of contemporary bioinformatics. A critical step of the process is to identify and retrieve relevant documents among the vast volume of modern biomedical literature efficiently. In the real world, many information retrieval tasks are difficult because of high data dimensionality and the lack of annotated examples to train a retrieval algorithm. Under such a scenario, the performance of information retrieval algorithms is often unsatisfactory, therefore improvements are needed. DESIGN: We studied two approaches that enhance the text categorization performance on sparse and high data dimensionality: (1) semantic-preserving dimension reduction by representing text with semantic-enriched features; and (2) augmenting training data with semi-supervised learning. A probabilistic topic model was applied to extract major semantic topics from a corpus of text of interest. The representation of documents was projected from the high-dimensional vocabulary space onto a semantic topic space with reduced dimensionality. A semi-supervised learning algorithm based on graph theory was applied to identify potential positive training cases, which were further used to augment training data. The effects of data transformation and augmentation on text categorization by support vector machine (SVM) were evaluated. RESULTS AND CONCLUSION: Semantic-enriched data transformation and the pseudo-positive-cases augmented training data enhance the efficiency and performance of text categorization by SVM.

Algorithms↗
PubMed2005

Activity of striatal neurons reflects dynamic encoding and recoding of procedural memories.

Learning to perform a behavioural procedure as a well-ingrained habit requires extensive repetition of the behavioural sequence, and learning not to perform such behaviours is notoriously difficult. Yet regaining a habit can occur quickly, with even one or a few exposures to cues previously triggering the behaviour. To identify neural mechanisms that might underlie such learning dynamics, we made long-term recordings from multiple neurons in the sensorimotor striatum, a basal ganglia structure implicated in habit formation, in rats successively trained on a reward-based procedural task, given extinction training and then given reacquisition training. The spike activity of striatal output neurons, nodal points in cortico-basal ganglia circuits, changed markedly across multiple dimensions during each of these phases of learning. First, new patterns of task-related ensemble firing successively formed, reversed and then re-emerged. Second, task-irrelevant firing was suppressed, then rebounded, and then was suppressed again. These changing spike activity patterns were highly correlated with changes in behavioural performance. We propose that these changes in task representation in cortico-basal ganglia circuits represent neural equivalents of the explore-exploit behaviour characteristic of habit learning.

Acoustic Stimulation↗
PubMed2002

Relearning of verbal labels in semantic dementia.

Semantic dementia is a degenerative disorder of temporal neocortex characterised by loss of word and object concepts. There is limited evidence that temporary relearning of lost vocabulary may be possible, attributed to sparing of hippocampal structures. However, learning is variable across patients and factors underlying learning success are poorly understood. The study investigated relearning of object names in two severely anomic semantic dementia patients. Following memory models that assume that hippocampal memories require some neocortical representation to underpin them it was predicted that relearning would be influenced by patients' residual semantic information about stimuli. Experiment 1 confirmed that residual knowledge influenced learning success. On the assumption that neocortical knowledge encompasses concepts of space and time, as well as words and objects, it was predicted that learning would be affected by the availability of contextual (temporo-spatial) information. Experiment 2 demonstrated effective learning of object names, attributed to the patient's use of temporal order and spatial position knowledge. Retention of object names over months was linked to the patient's capacity for autobiographical experiential (temporo-spatial contextual) association. The findings indicate that relearning of lost vocabulary is possible in semantic dementia, indicating a role of the medial temporal lobes in the acquisition of semantic information. Effective learning does not imply reinstatement of lost concepts, but, it is argued, does involve some reacquisition of meaning. The findings challenge the traditional semantic-episodic memory dichotomy and are consistent with a "levels of meaning" account of semantic memory.

Anomia↗
PubMed2005

Brain activation during execution and motor imagery of novel and skilled sequential hand movements.

This experiment used functional magnetic resonance imaging (fMRI) to compare functional neuroanatomy associated with executed and imagined hand movements in novel and skilled learning phases. We hypothesized that 1 week of intensive physical practice would strengthen the motor representation of a hand motor sequence and increase the similarity of functional neuroanatomy associated with executed and imagined hand movements. During fMRI scanning, a right-hand self-paced button press sequence was executed and imagined before (NOVEL) and after (SKILLED) 1 week of intensive physical practice (n = 54; right-hand dominant). The mean execution rate was significantly faster in the SKILLED (3.8 Hz) than the NOVEL condition (2.5 Hz) (P < 0.001), but there was no difference in execution errors. Activation foci associated with execution and imagery was congruent in both the NOVEL and SKILLED conditions, though activation features were more similar in the SKILLED versus NOVEL phase. In the NOVEL phase, activations were more extensive during execution than imagery in primary and secondary cortical motor volumes and the cerebellum, while during imagery activations were greater in the striatum. In the SKILLED phase, activation features within these same volumes became increasingly similar for execution and imagery, though imagery more heavily activated premotor areas, inferior parietal lobe, and medial temporal lobe, while execution more heavily activated the precentral/postcentral gyri, striatum, and cerebellum. This experiment demonstrated congruent activation of the cortical and subcortical motor system during both novel and skilled learning phases, supporting the effectiveness of motor imagery-based mental practice techniques for both the acquisition of new skills and the rehearsal of skilled movements.

Adult↗
PubMed1979

The anatomical substrate of callosal messages from SI and SII in the cat.

Horseradish peroxidase (HRP) was injected into the first (SI) or second (SII) somatosensory areas of 21 adult cats. The radial and tangential (normal and parallel to the pial surface, respectively) distribution and morphology of the callosal neurons were studied. HRP injections were combined with single unit recording in the contralateral cortex in order to determine which part of the somatosensory periphery is represented within the regions containing callosal neurons, the callosal (efferent) zones, in SI and SII. The callosal zone of SI extends over the trunk and part of the forepaw representation. In the forepaw and hindlimb representations callosal neurons projecting only to the contralateral SII are found, while in the trunk representation callosal neurons projecting to contralateral SI or SII are found. The callosal zone in SII extends widely throughout the forepaw representation in this area and projects to the contralateral SII but not to SI. In both SI and SII the callosal neurons are mainly located in layer III. A few of them are also found in layer VI. They are very rare in other layers. Callosal neurons in layer III are mostly pyramidal but exceptionally stellate; in layer VI they are pyramidal, triangular, and occasionally stellate. These data indicate that transformations of the cortical somatosensory maps are achieved in the message sent through the corpus callosum. These transformations are i) determined by the extent and location of the callosal zones and perhaps by the distribution of callosal neurons within them, ii) different in different areas, iii) different in a same area, according to the cortical targets to which they are conveyed. The existence of callosal connections originated from areas of distal forepaw representation supplies a possible anatomical substrate for those types of intermanual transfer of tactile learning which depend upon the integrity of the corpus callosum.

Animals↗
PubMed2000

Functional reorganisation of the corticomotor projection to the hand in skilled racquet players.

While it is known that relatively rapid changes in functional representation may occur in the human sensorimotor cortex in short-term motor-learning studies, there have been few studies of changes in organisation of the corticomotor system associated with the long-term acquisition of motor skills. In the present study, we have used transcranial magnetic stimulation (TMS) to investigate the corticomotor projection to the hand in a group of elite racquet players, who have developed and maintained a high level of skill over a period of many years, and have compared the findings with those in a group of social players and a group of non-playing control subjects. Increased motor-evoked-potential (MEP) amplitudes and shifts in the cortical motor maps for the playing hand were found in all of the elite players and cortical motor thresholds were reduced in some players, whereas in the social players all parameters were within the normal range. The findings in the elite players are interpreted as being indications of a process of functional reorganisation with the motor cortex or corticomotor pathway that are associated with the acquisition and retention of complex motor skills.

Adult↗
PubMed2006

Motor brain regions are involved in the encoding of delayed intentions: a fMRI study.

In studies of prospective memory, recall of the content of delayed intentions is normally excellent, probably because they contain actions that have to be enacted at a later time. Action words encoded for later enactment are more accessible from memory than those encoded for later verbal report [Freeman, J.E., and Ellis, J.A. 2003a. The representation of delayed intentions: A prospective subject-performed task? Journal of Experimental Psychology: Learning, Memory, and Cognition, 29, 976-992.]. As this higher assessibility is lost when the intended actions have to be enacted during encoding, or when a motor interference task is introduced concurrent to intention encoding, Freeman and Ellis suggested that the advantage of to-be-enacted actions is due to additional preparatory motor operations during encoding. Accordingly, in a fMRI study with 10 healthy young participants, we investigated whether motor brain regions are differentially activated during verbal encoding of actions for later enactment with the right hand in contrast to verbal encoding of actions for later verbal report. We included an additional condition of verbal encoding of abstract verbs for later verbal report to investigate whether the semantic motor information inherent in action verbs in contrast to abstract verbs activates motor brain regions different from those involved in the verbal encoding of actions for later enactment. Differential activation for the verbal encoding of to-be-enacted actions in contrast to to-be-reported actions was found in brain regions known to be involved in covert motor preparation for hand movements, i.e. the postcentral gyrus, the precuneus, the dorsal and ventral premotor cortex, the posterior middle temporal gyrus and the inferior parietal lobule. There was no overlap between these brain regions and those differentially activated during the verbal encoding of actions in contrast to abstract verbs for later verbal report. Consequently, the results of this fMRI study suggest the presence of preparatory motor operations during the encoding of delayed intentions requiring a future motor response, which cannot be attributed to semantic information inherent to action verbs.

Adult↗
PubMed2004

Developmental cognitive neuroscience: progress and potential.

Developmental cognitive neuroscience is an evolving field that investigates the relations between neural and cognitive development. Lying at the intersection of diverse disciplines, work in this area promises to shed light on classic developmental questions, mechanisms subserving developmental change, diagnosis and treatment of developmental disorders, and cognitive and neuroscientific topics traditionally considered outside the domain of development. Fundamental questions include: What are the interrelations between developmental changes in the brain (e.g. in connectivity, chemistry, morphology) and developmental changes in children's behavior and cognitive abilities (e.g. representational complexity, ability to sustain selective attention, speed of processing)? Why, and how, is learning enhanced during certain periods in development? How is our knowledge organized, and how does this change with development? We discuss preliminary investigations of such questions and directions for future work.

Aptitude↗
PubMed1999

Apraxia and motor-skill acquisition in Alzheimer's disease are dissociable.

Many patients with Alzheimer's disease (AD) are apraxic and the apraxia has been posited to be related to a loss of movement representations. Whereas patients with Alzheimer's disease have been reported to demonstrate normal motor learning on a rotor pursuit skill acquisition task, it is unknown whether AD subjects who are apraxic demonstrate normal skill-learning. We tested subjects with probable AD and normal controls on a rotor pursuit task. We also tested the AD subjects for ideomotor apraxia. Subjects with AD who were apraxic had normal motor learning. In addition, praxis score did not correlate with performance on the skill-acquisition task. The results suggest that ideomotor praxis and motor learning are at least partly dissociable.

Aged↗
PubMed2002

A network representation of response probability in the striatum.

The striatum of the basal ganglia is considered a key structure in the learning circuitry of the brain. To analyze neural signals that underlie striatal plasticity, we recorded from an identifiable class of striatal interneurons as macaque monkeys underwent training in a range of conditioning and non-associative learning paradigms, and recorded eyeblink electromyographs as the measure of behavioral response. We found that the responses of these striatal interneurons were modifiable under all training conditions and that their population responses were tightly correlated with the probability that a given stimulus would evoke a behavioral response. Such a network signal, proportional to current response probability, could be crucial to the learning and decision functions of the basal ganglia.

Animals↗
PubMed2002

Effects of early experience on children's recognition of facial displays of emotion.

The present research examines visual perception of emotion in both typical and atypical development. To examine the processes by which perceptual mechanisms become attuned to the contingencies of affective signals in the environment, the authors measured the sequential, content-based properties of feature detection in emotion recognition processes. To evaluate the role of experience, they compared typically developing children with physically abused children, who were presumed to have experienced high levels of threat and hostility. As predicted, physically abused children accurately identified facial displays of anger on the basis of less sensory input than did controls, which suggests that physically abused children have facilitated access to representations of anger. The findings are discussed in terms of experiential processes in perceptual learning.

Child↗
PubMed1995

Abstract visual-form representations in the left cerebral hemisphere.

Visual-form systems in the cerebral hemispheres were examined in 3 experiments. After learning new types of visual forms, participants rapidly classified previously unseen prototypes of the newly learned types more efficiently when the forms were presented directly to the left hemisphere (in the right visual field) than when the forms were presented directly to the right hemisphere (in the left visual field). Neither previously seen nor previously unseen distortions of the prototypes were classified more efficiently when presented directly to the left hemisphere than when presented directly to the right hemisphere. Results indicate that an abstract visual-form system operates effectively in the left hemisphere and stores information that remains relatively invariant across the specific instances of a type of form to distinguish different types. Furthermore, this system functions relatively independently of another system that operates effectively in the right hemisphere and that stores details to distinguish specific instances of a type of form.

Brain↗
PubMed2006

Leafing through the genomes of our major crop plants: strategies for capturing unique information.

Crop plants not only have economic significance, but also comprise important botanical models for evolution and development. This is reflected by the recent increase in the percentage of publicly available sequence data that are derived from angiosperms. Further genome sequencing of the major crop plants will offer new learning opportunities, but their large, repetitive, and often polyploid genomes present challenges. Reduced-representation approaches - such as EST sequencing, methyl filtration and Cot-based cloning and sequencing - provide increased efficiency in extracting key information from crop genomes without full-genome sequencing. Combining these methods with phylogenetically stratified sampling to allow comparative genomic approaches has the potential to further accelerate progress in angiosperm genomics.

Crops, Agricultural↗
PubMed1994

The Influence of Skill and Intermittent Vision on Dynamic Balance.

Two experiments are reported in which expert and novice gymnasts were required to walk across a balance beam as quickly as possible in various vision conditions. In Experiment 1, experts walked faster than novices in all vision conditions, showing the greatest superiority when vision was completely eliminated. Novices were more dependent on vision and were able to maintain their performance as long as a visual sample was available every 250 ms (i.e: 4-Hz samples).The results of Experiment 2 indicate that differences between expert and novice performers in the no-vision condition were not related to the use of a short-term visual representation of the movement environment. Our movement time findings are problematic for specificity of learning models of skill acquisition. As well, film data collected in Experiment 2 were not consistent with models that propose a transition from closed-loop to open-loop control.

Journal Article↗
PubMed2002

Prefrontal cortex and dynamic categorization tasks: representational organization and neuromodulatory control.

We present a computational model of the intradimensional/ extradimensional (ID/ED) task (a variant of the Wisconsin card sorting task) that simulates the performance of intact and frontally lesioned monkeys on three different kinds of rule changes (Dias et al., 1997, J Neurosci 17:9285-9297). Although Dias et al. interpret the lesion data as supporting a model in which prefrontal cortex is organized into different processing functions, our model suggests an alternative account based on representational content. A key aspect of the model is that prefrontal cortex representations are organized according to different levels of abstraction, with orbital areas encoding more specific featural information and dorsolateral areas encoding more abstract dimensional information. This representational scheme of the model is integrated with two additional key elements: (i) activation-based working memory representations controlled by a dynamic gating mechanism that simulates the hypothesized phasic actions of dopaminergic neuromodulation in prefrontal cortex, which acts to stabilize or destabilize frontal representations based on success in the task; and (ii) a weight-based associative learning system simulating posterior cortex and other subcortical areas, where the stimulus-response mappings are encoded. Frontal cortex contributes to the task via top-down activation-based biasing of task-appropriate features and dimensions in this posterior cortex system - this top-down biasing is specifically important for overcoming prepotent associations after a sorting rule reverses. The ability of the model to capture the double-dissociation observed by Dias et al. with orbital versus dorsolateral lesions supports the validity of these principles, many of which have also been useful in accounting for other frontal phenomena.

Animals↗
PubMed2005

Structural semantic interconnections: a knowledge-based approach to word sense disambiguation.

Word Sense Disambiguation (WSD) is traditionally considered an Al-hard problem. A break-through in this field would have a significant impact on many relevant Web-based applications, such as Web information retrieval, improved access to Web services, information extraction, etc. Early approaches to WSD, based on knowledge representation techniques, have been replaced in the past few years by more robust machine learning and statistical techniques. The results of recent comparative evaluations of WSD systems, however, show that these methods have inherent limitations. On the other hand, the increasing availability of large-scale, rich lexical knowledge resources seems to provide new challenges to knowledge-based approaches. In this paper, we present a method, called structural semantic interconnections (SSI), which creates structural specifications of the possible senses for each word in a context and selects the best hypothesis according to a grammar G, describing relations between sense specifications. Sense specifications are created from several available lexical resources that we integrated in part manually, in part with the help of automatic procedures. The SSI algorithm has been applied to different semantic disambiguation problems, like automatic ontology population, disambiguation of sentences in generic texts, disambiguation of words in glossary definitions. Evaluation experiments have been performed on specific knowledge domains (e.g., tourism, computer networks, enterprise interoperability), as well as on standard disambiguation test sets.

Algorithms↗