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 415 records · Page 23Linked to original sources
PubMed2004

Knowing where things are in the second year of life: implications for hippocampal development.

Prior data have revealed striking contrasts between 18- and 24-month-old children in place learning, an ability known to depend on the hippocampus (Newcombe, Huttenlocher, Drummey, & Wiley, 1998). The current research examined the development of three other basic abilities of mature spatial competence: the representation of multiple locations, the learning of relations among objects, and the recall of a single location after a substantial filled delay. Results indicated a transition from 18 to 24 months in all three abilities. This evidence supports a general transition in spatial representation that occurs towards the end of infancy. Existing neurobehavioral data suggest that a corresponding change in hippocampal functioning underlies this development.

Age Factors↗
PubMed2005

The generation of conscious awareness in an incidental learning situation.

The purpose of the present experiments was to investigate the generation of conscious awareness (i.e., of verbal report) in an incidental learning situation. While the single-system account assumes that all markers of learning, verbal or nonverbal, index the same underlying knowledge representation, multiple-systems accounts grant verbal report a special status as a marker of learning because they assume that the nonverbal and verbal effects of learning rely on different memory representations. We tested these two accounts in two experiments in which we held the amount of learning in the nonverbal memory system constant while manipulating independent variables aimed at affecting learning in the declarative system. The results of both experiments revealed significant differences in verbal report between experimental conditions, but no significant differences in response times. Overall, these results provide clear evidence in favor of the multiple-systems account.

Adolescent↗
PubMed2002

Posttraining glucocorticoid receptor agonist enhances memory in appetitive and aversive Pavlovian discrete-cue conditioning paradigms.

Glucocorticoid modulation of emotional memory has repeatedly been shown in aversive learning paradigms, but has received little attention in appetitive tasks. It has also been suggested that it may be selective for contextual cues. In order to investigate if glucocorticoids can modulate memory in discrete-cue conditioning of both appetitive and aversive tasks, two experiments were carried out. Male Lister-Hooded rats received pairings of an auditory cue and either food-reward (experiment 1) or footshock (experiment 2), followed immediately by posttraining injections of the glucocorticoid receptor agonist dexamethasone (0.3, 0.6, and 1.2 mg/kg) or vehicle. Dexamethasone (1.2 mg/kg) led to significantly enhanced learning. These results give support to the notion that glucocorticoids play a role in the modulation of both appetitive and aversive emotional memories and show that their role in learning goes beyond the construction of context representations. The modulation of appetitive and aversive discrete-cue learning may be subserved by a common mechanism.

Animals↗
PubMed2005

Non-symbolic arithmetic in adults and young children.

Five experiments investigated whether adults and preschool children can perform simple arithmetic calculations on non-symbolic numerosities. Previous research has demonstrated that human adults, human infants, and non-human animals can process numerical quantities through approximate representations of their magnitudes. Here we consider whether these non-symbolic numerical representations might serve as a building block of uniquely human, learned mathematics. Both adults and children with no training in arithmetic successfully performed approximate arithmetic on large sets of elements. Success at these tasks did not depend on non-numerical continuous quantities, modality-specific quantity information, the adoption of alternative non-arithmetic strategies, or learned symbolic arithmetic knowledge. Abstract numerical quantity representations therefore are computationally functional and may provide a foundation for formal mathematics.

Adult↗
PubMed1997

Increased Sensitivity to Changes in Visual Feedback With Practice.

In recent work investigating motor learning, the focus has been on the effect of modifying feedback at different levels of learning. Results suggest that learning is specific to the practiced conditions and that this specificity increases with practice. In a replication and extension of this previous work, 3 groups (N = 30 subjects) practiced a sequential positioning movement: Controls performed 300 trials with visually presented on-line kinematic feedback, whereas the other 2 groups, low practice (LP) or high practice (HP), performed, respectively, 50 or 300 trials without feedback. Pretest and posttest sessions of 10 trials each were performed with the on-line feedback. All groups improved with practice. It was apparent that the HP group exhibited more of a performance decrement in the posttest than the LP group did, suggesting that motor learning is the process of forming an increasingly specific sensorimotor representation. These results have implications for motor learning paradigms, models of motor learning, and training.

humans↗
PubMed1998

Rules and exemplars in category learning.

Psychological theories of categorization generally focus on either rule- or exemplar-based explanations. We present 2 experiments that show evidence of both rule induction and exemplar encoding as well as a connectionist model, ATRIUM, that specifies a mechanism for combining rule- and exemplar-based representation. In 2 experiments participants learned to classify items, most of which followed a simple rule, although there were a few frequently occurring exceptions. Experiment 1 examined how people extrapolate beyond the range of training. Experiment 2 examined the effect of instance frequency on generalization. Categorization behavior was well described by the model, in which exemplar representation is used for both rule and exception processing. A key element in correctly modeling these results was capturing the interaction between the rule- and exemplar-based representations by using shifts of attention between rules and exemplars.

Attention↗
PubMedDate not supplied

New approaches to the study of amnesic patients: what can a neurofunctional philosophy and neural network methods offer?

In this paper I first consider a neurofunctional approach to the study of amnesic patients. This approach stresses the need for theorising about the processing operations of brain regions and circuits rather than for theorising about neuropsychological syndromes. A syndrome such as amnesia-may not exist, in any meaningful sense, if there is marked heterogeneity within the patients grouped together in this way. Powerful neuroimaging techniques may now allow a more useful basis for grouping patients in terms of lesion location rather than aetiology. In turn this will allow an evaluation of the information processing functions subserved by the lesioned structures. The second strand to the present paper stresses the weakness in the specification of current theories. This has made it difficult to select experimental tasks that decisively measure the key components of those theories. The paper makes the case that explicit neural network models are a useful way to try to overcome this problem. In line with these ideas, the paper begins to build a model of how the brain may achieve useful kinds of stimulus representations. Considerations of human behaviour in category learning tasks have emphasised parallel and interacting roles for both exemplar- and element-based stimulus representations. It is suggested that the hippocampus itself may encode exemplar representations, and these may provide a basis for episodic memory as well as some types of category learning. It is further suggested that the ventral striatum may encode the element-based representations. The model allows some new and detailed predictions for the performance of amnesic subjects related to lesion location.

Amnesia↗
PubMed2001

Visual perceptual learning in human object recognition areas: a repetition priming study using high-density electrical mapping.

It is often the case that only partial or degraded views of an object are available to an observer, and yet in many of these cases, object recognition is accomplished with surprising ease. The perceptual filling-in or "closure" that makes this possible has been linked to a group of object recognition areas in the human brain, the lateral occipital (LO) complex, and has been shown to have a specific electrophysiological correlate, the N(cl) component of the event related potential. Perceptual closure presumably occurs because repeated and varied exposure to different classes of objects has caused the brain to undergo "perceptual learning," which promotes a robust mnemonic representation, accessible under partial information circumstances. The present study examined the impact of perceptual learning on closure-related brain processes. Fragmented pictures of common objects were presented, such that information content was incrementally increased until just enough information was present to permit closure and object recognition. Periodic repetition of a subset of these picture sequences was used to induce repetition priming due to perceptual learning. This priming has an electrophysiological signature that is putatively generated in the LO complex, but significantly precedes the electrophysiological correlate of closure. The temporal progression of priming- and closure-related activity in the LO complex supports the view that sensory processing entails multiple reentrant stages of activity within processing modules of the visual hierarchy. That the earliest priming-related activity occurs over LO complex, suggests that the sensory trace itself may reside in these object recognition areas.

Adult↗
PubMed2002

Dynamic representational plasticity in sensory cortex.

Studies of the effects of peripheral and central lesions, perceptual learning and neurochemical modification on the sensory representations in cortex have had a dramatic effect in alerting neuroscientists and therapists to the reorganizational capacity of the adult brain. An intriguing aspect of some of these investigations, such as partial peripheral denervation, is the short-term expression of these changes. Indeed, in visual cortex, auditory cortex and somatosensory cortex loss of input from a region of the peripheral receptor epithelium (retinal, basilar and cutaneous, respectively) induces rapid expression of ectopic, or expanded, receptive fields of affected neurons and reorganization of topographic maps to fill in the representation of the denervated area. The extent of these changes can, in some cases, match the maximal extents demonstrated with chronic manipulations. The rapidity, and reversibility, of the effects rules out many possible explanations which involve synaptic plasticity and points to a capacity for representational plasticity being inherent in the circuitry of a topographic pathway. Consequently, topographic representations must be considered as manifestations of physiological interaction rather than as anatomical constructs. Interference with this interaction can produce an unmasking of previously inhibited responsiveness. Consideration of the nature of masking inhibition which is consistent with the precision and order of a topographic representation and which has a capacity for rapid plasticity requires, in addition to stimulus-driven inhibition, a source of tonic input from the periphery. Such input, acting locally to provide tonic inhibition, has been directly demonstrated in the somatosensory system and is consistent with results obtained in auditory and visual systems.

Animals↗
PubMed2001

Prototype formation in autism.

Individuals with autism have difficulty integrating information and generalizing previously learned concepts to new situations. It was hypothesized that these problems result from an underlying impairment in category formation. Persons with autism may not abstract a summary representation (a prototype) during category learning and, instead, may form categories by memorizing a list of rules. Children with autism, Down syndrome, and normal development participated in one set of category learning tasks that could be solved using a rule-based approach and a second set of tasks in which there was no rule that defined category membership (prototype tasks). In the rule-based tasks, all groups were successful at using a rule to learn a new category. In the prototype tasks, only the typically developing children were able to learn a new category. Neither the persons with autism nor the persons with Down syndrome appeared to develop a prototype during category learning. These data suggest that persons with autism and Down syndrome have difficulty categorizing new information by forming prototypes and, instead, tend to rely on a rule-based approach to learning.

Adolescent↗
PubMed2004

D-cycloserine and the facilitation of extinction of conditioned fear: consequences for reinstatement.

Several recent studies have reported that D-cycloserine (DCS), a partial N-methyl-D-aspartate agonist, facilitates extinction of learned fear in rats. Other studies have shown that representation of the unconditioned stimulus (US) can reinstate learned fear after extinction. This study examined whether this reinstatement effect occurs in Sprague-Dawley rats given DCS at the time of extinction. Results showed that saline-treated rats exhibited the reinstatement effect but DCS-treated rats did not (Experiments 1 and 2). This lack of reinstatement in DCS-treated rats was not due to residual effects of DCS on either US or context processing (Experiment 3). Overall, these results (a) raise questions about the mechanisms underlying DCS facilitation of extinction and (b) suggest that DCS might have substantial practical benefit.

Analysis of Variance↗
PubMed2005

Novices in clinical practice settings: student nurses stories of learning the practice of nursing.

Drawing on 24 stories of clinical practice in an apprenticeship context of training in Israel, this qualitative study examined student nurses' perspectives towards learning to become a nurse, as revealed through the language and content of their written stories of clinical practice. As our findings suggest, student nurses' stories of learning to become a nurse in practice settings, are characterized by procedural language, by medical rather than nursing terminology, and by a focus on actions rather than on interactions. We have learned that, despite the rich content that characterizes clinical practice settings, the apprenticeship orientation of the training program, combined with student nurses' state of being a novice, yielded representations of the experience of learning to nurse which were characterized by an instrumental perspective towards the practice. We interpret these findings through four interrelated insights that emerge from the study: (1) an 'instrumental practice' orientation in the setting of caring, (2) knowledge of clinical facts-not knowledge of clinical principles, (3) the fragmented character of novices' learning to nurse in practice, and (4) rich content of practice alone does not yield rich content of learning.

Adaptation, Psychological↗
PubMed1984

Induction of category distributions: a framework for classification learning.

We present a framework for classification learning that assumes that learners use presented instances (whether labeled or unlabeled) to infer the density functions of category exemplars over a feature space and that subsequent classification decisions employ a relative likelihood decision rule based on these inferred density functions. A specific model based on this general framework, the category density model, was proposed to account for the induction of normally distributed categories either with or without error correction or provision of labeled instances. The model was implemented as a computer simulation. Results of five experiments indicated that people could learn category distributions not only without error correction, but without knowledge of the number of categories or even that there were categories to be learned. These and other findings dictated a more general learning model that integrated distributional representations based on both parametric descriptions and stored instances.

Decision Making↗
PubMed2025

scMGCL: accurate and efficient integration representation of single-cell multi-omics data.

MOTIVATION: Single-cell multi-omics data integration is essential for understanding cellular states and disease mechanisms, yet integrating heterogeneous data modalities remains a challenge. We present scMGCL, a graph contrastive learning framework for robust integration of single-cell ATAC-seq and RNA-seq data. Our approach leverages self-supervised learning on cell-cell similarity graphs, in which each modality's graph structure serves as an augmentation for the other. This cross-modality contrastive paradigm enables the learning of biologically meaningful, shared representations while preserving modality-specific features. RESULTS: Benchmarking against state-of-the-art methods demonstrates that scMGCL outperforms others in cell-type clustering, label transfer accuracy, and preservation of marker-gene correlations. Additionally, scMGCL significantly improves computational efficiency, reducing runtime and memory usage. The method's effectiveness is further validated through extensive analyses of cell-type similarity and functional consistency, providing a powerful tool for multi-omics data exploration. AVAILABILITY AND IMPLEMENTATION: Code and datasets are released at https://github.com/zlCreator/scMGCL.

Single-Cell Analysis↗
PubMed2001

Best harmony, unified RPCL and automated model selection for unsupervised and supervised learning on Gaussian mixtures, three-layer nets and ME-RBF-SVM models.

After introducing the fundamentals of BYY system and harmony learning, which has been developed in past several years as a unified statistical framework for parameter learning, regularization and model selection, we systematically discuss this BYY harmony learning on systems with discrete inner-representations. First, we shown that one special case leads to unsupervised learning on Gaussian mixture. We show how harmony learning not only leads us to the EM algorithm for maximum likelihood (ML) learning and the corresponding extended KMEAN algorithms for Mahalanobis clustering with criteria for selecting the number of Gaussians or clusters, but also provides us two new regularization techniques and a unified scheme that includes the previous rival penalized competitive learning (RPCL) as well as its various variants and extensions that performs model selection automatically during parameter learning. Moreover, as a by-product, we also get a new approach for determining a set of 'supporting vectors' for Parzen window density estimation. Second, we shown that other special cases lead to three typical supervised learning models with several new results. On three layer net, we get (i) a new regularized ML learning, (ii) a new criterion for selecting the number of hidden units, and (iii) a family of EM-like algorithms that combines harmony learning with new techniques of regularization. On the original and alternative models of mixture-of-expert (ME) as well as radial basis function (RBF) nets, we get not only a new type of criteria for selecting the number of experts or basis functions but also a new type of the EM-like algorithms that combines regularization techniques and RPCL learning for parameter learning with either least complexity nature on the original ME model or automated model selection on the alternative ME model and RBF nets. Moreover, all the results for the alternative ME model are also applied to other two popular nonparametric statistical approaches, namely kernel regression and supporting vector machine. Particularly, not only we get an easily implemented approach for determining the smoothing parameter in kernel regression, but also we get an alternative approach for deciding the set of supporting vectors in supporting vector machine.

Algorithms↗
PubMed2004

[Educating the patient: skills needed to nursing at the crossroads of care and pedagogy].

The educational team interrogated itself concerning its pedagogical competence within the framework of its educational activity for asthmatics: how does the care provider become a learning facilitator for the asthmatic person? Among the possible responses, there were two aspects which appeared essential: understanding the patient's learning logic and taking into consideration his/her representations and prior knowledge. The concrete translation of these principles primarily rests upon an educational programme truly centred on the person in question, and one which uses interactive methods, promoting the transfer of knowledge in daily life. Then the caregiver utilises previous experience and the asthmatic's representations in order to solicit a cognitive conflict, a source of learning.

Asthma↗
PubMed2004

Adult age differences in self-regulated learning from reading sentences.

We examined age differences in the heuristic used to allocate effort in learning information from sentences. Younger and older adults read and reread sentences varying in propositional density for recall, making judgments of learning before producing recall. The allocation of effort in rereading items that were less well learned on the first reading was optimized for sentences of intermediate complexity, especially for older adults. These data support a model of self-regulated learning in which readers reduce the discrepancy between current and optimal states of learning. However, self-regulation, which may be procedure based or rely on an implicit representation of the current state of learning, may be particularly efficient for older adults within a region of proximal learning.

Adult↗
PubMed2003

Definition of a new maze paradigm for the study of spatial behavior in rats.

The present work defines a simple behavioral paradigm to evaluate spatial representation in rats. In two experimental conditions differing in the richness of distal visual cues, rats learned to locate a food goal in a cross maze from various starting points. We conducted different challenges consisting of (i) reaching the same goal from a modified arrangement of the maze arms (geometric challenge), (ii) reaching the goal within a 90 degrees rotated maze, herein checking the use of a place strategy, and (iii) investigating the effect of central cholinergic blockade over the retention of the task. Results showed that rats needed 12-30 trials to learn a place response, depending upon the richness of the visual environment. The maze rotation did not produce any impairment whereas the geometric challenge affected the performance specifically under the visually richer environment. Scopolamine injection (i.p.) produced a significant impairment in place recognition. Our present work shows that this maze procedure constitutes a useful paradigm to assess learning and processing of a place representation by rats. Similarly to what has been shown in other popular maze paradigms, our results show that rats mostly rely on distal extra-maze cues to solve the task, but also compute intra-maze information.

Animals↗