On the internal representation of associative data network.
Explore the source record for details and available documents.
Female↗
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.
[Comparative experiments on multidimensional scaling of subjective color differences and their internal spherical representation].
Explore the source record for details and available documents.
Color Perception↗
View dependence in scene recognition after active learning.
Human spatial encoding of three-dimensional navigable space was studied, using a virtual environment simulation. This allowed subjects to become familiar with a realistic scene by making simulated rotational and translational movements during training. Subsequent tests determined whether subjects could generalize their recognition ability by identifying novel-perspective views and topographic floor plans of the scene. Results from picture recognition tests showed that familiar direction views were most easily recognized, although significant generalization to novel views was observed. Topographic floor plans were also easily identified. In further experiments, novel-view performance diminished when active training was replaced by passive viewing of static images of the scene. However, the ability to make self-initiated movements, as opposed to watching dynamic movie sequences, had no effect on performance. These results suggest that representation of navigable space is view dependent and highlight the importance of spatial-temporal continuity during learning.
Adult↗
Functional measurements of human ventral occipital cortex: retinotopy and colour.
Human colour vision originates in the cone photoreceptors, whose spatial density peaks in the fovea and declines rapidly into the periphery. For this reason, one expects to find a large representation of the cone-rich fovea in those cortical locations that support colour perception. Human occipital cortex contains several distinct foveal representations including at least two that extend onto the ventral surface: a region thought to be critical for colour vision. To learn more about these ventral signals, we used functional magnetic resonance imaging to identify visual field maps and colour responsivity on the ventral surface. We found a visual map of the complete contralateral hemifield in a 4 cm(2) region adjacent to ventral V3; the foveal representation of this map is confluent with that of areas V1/2/3. Additionally, a distinct foveal representation is present on the ventral surface situated 3-5 cm anterior from the confluent V1/2/3 foveal representations. This organization is not consistent with the definition of area V8, which assumes the presence of a quarter field representation adjacent to V3v. Comparisons of responses to luminance-matched coloured and achromatic patterns show increased activity to the coloured stimuli beginning in area V1 and extending through the new hemifield representation and further anterior in the ventral occipital lobe.
Brain Mapping↗
A neuro-fuzzy framework for inferencing.
Earlier we proposed a connectionist implementation of compositional rule of inference (COI) for rules with antecedents having a single clause. We first review this net, then generalize it so that it can deal with rules with antecedent having multiple clauses. We call it COIN, the compositional rule of inferencing network. Given a relational representation of a set of rules, the proposed architecture can realize the COI. The outcome of COI depends on the choice of both the implication function and the inferencing scheme. The problem of choosing an appropriate implication function is avoided through neural learning. COIN can automatically find a 'good' relation to represent a set of fuzzy rules. We model the connection weights so as to ensure learned weights lie in [0,1]. We demonstrate through extensive numerical examples that the proposed neural realization can find a much better representation of the rules than that by usual implication and hence results in much better conclusions than the usual COI.
Fuzzy Logic↗
Evaluation and development of a connectionist theory of configural learning.
A configural theory of associative learning is described that is based on the assumption that conditioning results in associations between the unconditioned stimulus and a representation of the entire pattern of stimulation that was present prior to its delivery. Configural theory was formulated originally to account for generalization and discrimination in Pavlovian conditioning. The first part of the article demonstrates how this theory can be used to explain results from studies of overshadowing, blocking, summation, and discrimination learning. The second part of the article shows how the theory can be developed to explain a broader range of phenomena, including mediated conditioning, reinforcer devaluation effects, the differential outcomes effect, acquired equivalence, sensory preconditioning, and structural discriminations.
Animals↗
A critical evaluation of the force control hypothesis in motor control.
The ability to formulate explicit mathematical models of motor systems has played a central role in recent progress in motor control research. As a result of these modeling efforts and in particular the incorporation of concepts drawn from control systems theory, ideas about motor control have changed substantially. There is growing emphasis on motor learning and particularly on predictive or anticipatory aspects of control that are related to the neural representation of dynamics. Two ideas have become increasingly prominent in mathematical modeling of motor function--forward internal models and inverse dynamics. The notion of forward internal models which has drawn from work in adaptive control arises from the recognition that the nervous system takes account of dynamics in motion planning. Inverse dynamics, a complementary way of adjusting control signals to deal with dynamics, has proved a simple means to establish the joint torques necessary to produce desired movements. In this paper, we review the force control formulation in which inverse dynamics and forward internal models play a central role. We present evidence in its favor and describe its limitations. We note that inverse dynamics and forward models are potential solutions to general problems in motor control--how the nervous system establishes a mapping between desired movements and associated control signals, and how control signals are adjusted in the context of motor learning, dynamics and loads. However, we find little empirical evidence that specifically supports the inverse dynamics or forward internal model proposals per se. We further conclude that the central idea of the force control hypothesis--that control levels operate through the central specification of forces--is flawed. This is specifically evident in the context of attempts to incorporate physiologically realistic muscle and reflex mechanisms into the force control model. In particular, the formulation offers no means to shift between postures without triggering resistance due to postural stabilizing mechanisms.
Animals↗
Functional brain imaging studies of cortical mechanisms for memory.
Recent functional brain imaging studies in humans indicate that learning and memory involve many of the same regions of the cortex that process sensory information and control motor output. The forms of perceptual and motor learning that can occur without conscious recollection are mediated in part by contractions and expansions of representations in the sensory and motor cortex. The same regions are also engaged during the conscious storage and retrieval of facts and events, but these types of memory also bring into play structures involved in the active maintenance of memories "on line" and in the establishment of associative links between the information stored in different sensory areas. Although the picture of memory that is emerging from functional imaging studies is consistent with current physiological accounts, there are puzzles and surprises that will be solved only through a combination of human and animal studies.
Animals↗
Long-term potentiation induces expanded movement representations and dendritic hypertrophy in layer V of rat sensorimotor neocortex.
While long-term potentiation (LTP) is currently the most widely investigated model of the synaptic mechanisms underlying learning, there is a paucity of reports on the direct effects of LTP on cortical organization. Here we show that strengthening polysynaptic potentiation correlates with an expanded neocortical area that responds to intracortical microstimulation-induced movements of rat forelimb and increased dendritic material in layer V pyramidal cells. Rats carried a stimulating electrode in the corpus callosum (midline), and a recording electrode in the right caudal forelimb area (CFA). Each rat received 15 days of either high frequency stimulation (HFS) or handling. Evoked potentials of the transcallosal pathway were recorded in the right hemisphere before and after 15 days of stimulation or handling. Following the last stimulation, movement representations were determined in the left CFA using high-resolution intracortical microstimulation (ICMS) and then the brains were processed for Golgi-Cox staining. Our results show that synaptic modification results in a recruitment of more neocortical area into movement representations and increases in several measures of dendritic morphology in layers III and V. This study sheds light on the interaction between artificial models of learning, receptive field characteristics and dendritic morphology in the sensorimotor cortex.
Adaptation, Physiological↗
Complexity effects in visuo-spatial working memory: implications for the role of long-term memory.
Several studies have shown that the capacity of visuo-spatial working memory is limited by complexity. Using a variant of the Corsi blocks task, this paper investigates the effect of complexity of the to-be-remembered path on visuo-spatial memory span. Redundancy was determined by three Gestalt principles: symmetry, repetition, and continuation. Experiment 1 revealed an effect of path complexity. The subsequent experiments explored whether the superiority for recall of structured over complex paths can be attributed solely to the operation of visuo-spatial working memory, or whether it also reflects the use of long-term knowledge. Experiment 2 demonstrated that the effect of complexity remained, even when the mechanisms for visuo-spatial coding were removed by a secondary visuo-spatial task. In Experiments 3 and 4 subjects were trained in the recall of complex paths. This led to the creation of long-term memory representations for these paths, as shown by an improvement in their span, and a concomitant lack of transfer to new paths. Finally, Experiment 5 showed that one prior repetition of a complex path was sufficient to produce specific and long-term learning effects. These results point to the involvement of long-term memory processes in the temporary retention of visuo-spatial material for which representations exist in long-term memory. They also suggest that the effect of complexity may provide a tractable technique for investigating the mechanisms underlying the limits of visuo-spatial short-term storage.
Humans↗
Hebbian imprinting and retrieval in oscillatory neural networks.
We introduce a model of generalized Hebbian learning and retrieval in oscillatory neural networks modeling cortical areas such as hippocampus and olfactory cortex. Recent experiments have shown that synaptic plasticity depends on spike timing, especially on synapses from excitatory pyramidal cells, in hippocampus, and in sensory and cerebellar cortex. Here we study how such plasticity can be used to form memories and input representations when the neural dynamics are oscillatory, as is common in the brain (particularly in the hippocampus and olfactory cortex). Learning is assumed to occur in a phase of neural plasticity, in which the network is clamped to external teaching signals. By suitable manipulation of the nonlinearity of the neurons or the oscillation frequencies during learning, the model can be made, in a retrieval phase, either to categorize new inputs or to map them, in a continuous fashion, onto the space spanned by the imprinted patterns. We identify the first of these possibilities with the function of olfactory cortex and the second with the observed response characteristics of place cells in hippocampus. We investigate both kinds of networks analytically and by computer simulations, and we link the models with experimental findings, exploring, in particular, how the spike timing dependence of the synaptic plasticity constrains the computational function of the network and vice versa.
Hippocampus↗
Evidence for abstract, schematic knowledge of three spatial diagram representations.
Spatial diagram representations such as hierarchies, matrices, and networks are important tools for thinking. Our data suggest that college students possess abstract schemas for these representations that include at least rudimentary information about their applicability conditions. In Experiment 1, subjects were better able to select the appropriate spatial diagram representation for a problem when cued to use general category information in memory about those representations than when cued to use specific example problems given during the experiment. The results of Experiment 2 showed that the superior performance in the general category condition was not based on a comparison of the test problems with examples in memory. The results of Experiment 3 showed that the superior performance was not due to learning that occurred during the experiment or to transfer appropriate processing. The General Discussion section considers the nature of students' representation schemas and the question of why college students have only rudimentary schemas for common and widely applicable diagrammatic representations.
Female↗
Role of dorsomedial thalamic nucleus and piriform cortex in processing olfactory information.
Four experiments were conducted to characterize the role of primary and secondary olfactory projection areas (piriform cortex and dorsomedial thalamic nucleus (DMN] in olfactory information processing. Rats had to learn to discriminate between odors that were simultaneously released from different arms of an automated olfactory maze. When standard training conditions were used, damage of the DMN severely impaired both preoperatively trained and naive animals in acquiring an odor discrimination set (i.e. in most problems no learning was demonstrated). An additional group of DMN animals that received 4 times the standard amount of daily trials was unable to acquire the first two problems but successfully solved the third and all subsequent discriminations. Analysis of performance patterns suggested that destruction of the DMN initially leads to a strong procedural impairment that can be overcome by extensive training. After solving the third problem the animals with DMN damage required much less training to reach the learning criterion but generally made more errors than controls. Transfer of savings rarely occurred when a problem was repeated. Whether this secondary learning deficit observed in later discriminations is due to a specific effect of the lesion on the encoding of olfactory cues and thus on memory formation, or due to a disturbance in the regulation of emotional factors such as motivation, arousal, and attention is discussed. Lesions of the thalamus that spared the DMN had no effect on learning or retention of olfactory discriminations. Animals with ablations of the piriform cortex only acquired odor discriminations if they had been trained in the olfactory maze before the lesion. Moreover, their performance depended on the odor quality: they had great difficulty learning complex cues consisting of several odorants and learned simple odors virtually identical to control rats. The results indicate that an intact piriform cortex is needed to acquire the procedures involved to perform an olfactory discrimination task as well as to build neural representations of olfactory cues.
Animals↗
Recognizing biothreat diseases: realistic training using standardized patients and patient simulators.
The UofL Center for the Deterrence of Biowarfare and Bioterrorism's (CDBB's) training involving standardized patients, who can make convincingly accurate representations of clinical signs and symptoms, and patient simulators has proven to be an effective means of enhancing bioterrorism preparedness. In addition to providing ready access to formative measures of preparedness, both of these teaching and learning tools also offer the option of summative evaluation of skills and knowledge acquired during training sessions. The use of moulage allows for very realistic representations of cutaneous anthrax and smallpox as well as other conditions such as recluse spider bites and chickenpox with which these biothreat infections can be confused. Exercises and drills expose deficiencies and permit essential reinforcement of skills developed in training. Unannounced exercises are especially useful in assessing public health preparedness.
Biological Warfare↗
Learning and transfer of object-reward associations and the role of the perirhinal cortex.
Perirhinal cortex ablation has previously been shown only to impair new postoperative object discrimination learning with large stimulus set sizes (> or = 40 problems). In this study, 3 cynomolgus monkeys (Macaca fascicularis) with bilateral perirhinal cortex ablations were impaired relative to 3 normal controls on concurrent discrimination learning tasks with only 10 problems with the objects presented in different orientations in each trial to increase the demands placed on object identification. This supports the hypothesis that perirhinal cortex damage impairs the ability to identify multiple individual objects. Fewer errors were made to digitized images of objects than toward real objects. Both groups subsequently transferred specific object-reward associations from real objects to digitized images of the respective objects and vice versa, providing evidence that cynomolgus monkeys can recognize photographic representations of objects with experience.
Animals↗
Level and mechanisms of perceptual learning: learning first-order luminance and second-order texture objects.
Perceptual learning is an improvement in perceptual task performance reflecting plasticity in the perceptual system. Practice effects were studied in two object orientation tasks: a first order, luminance object task and a second-order, texture object task. Perceptual learning was small or absent in the first-order task, but consistently occurred for the second-order (texture) task, where it was limited to improvements in low external noise conditions, or stimulus enhancement [Dosher, B., & Lu, Z. -L. (1998). Perceptual learning reflects external noise filtering and internal noise reduction through channel reweighting. Proceedings of the National Academy of Sciences of the United States of America, 95 (23) 13988-13993; Dosher, B., & Lu, Z. -L. (1999). Mechanisms of perceptual learning. Vision Research, 39 (19) 3197-3221], analogous to attention effects in first- and second-order motion processing [Lu, Z. -L., Liu, C. Q., & Dosher, B. (2000). Attention mechanisms for multi-location first- and second-order motion perception. Vision Research, 40 (2) 173-186]. Perceptual learning affected the later, post-rectification, stages of perceptual analysis, possibly localized at V2 or above. It serves to amplify the stimulus relative to limiting internal noise for intrinsically noisy representations of second-order stimuli.
Attention↗
Partial blocking of NMDA receptors reduces plastic changes induced by short-lasting classical conditioning in the SI barrel cortex of adult mice.
The effect of blockade of N-methyl-D-aspartate (NMDA) receptors in the barrel cortex upon the learning-induced changes of the cortical body map was examined in adult mice. We have previously found that three sensory conditioning sessions, in which stimulation of a row of vibrissae was paired with a tail shock, produced an enlargement of the functional representation of a row of vibrissae stimulated during training. Implantation of the slow release polymer Elvax, containing 2-amino-5-phosphonovalerate (APV, 50 mM), in the vicinity of the barrel cortex was performed 1 day before conditioning to block NMDA receptors. The cortical representation of a trained row of vibrissae was visualized with 2-deoxyglucose (2DG) functional brain mapping 1 day after the completion of the conditioning procedure. The partial blockade of NMDA receptors within the barrel cortex reduced (by half) the expansion of the cortical representation of a trained row of vibrissae as compared to the enlargement of the cortical representation of a trained row found in untreated (60%) and Elvax-PBS implanted (47%) mice. The results provide evidence that the learning-induced processes of cortical map reorganization involve mechanisms that depend on NMDA receptor activation.
2-Amino-5-phosphonovalerate↗
A statistical basis for speech sound discrimination.
Infants under six months are able to discriminate native and non-native consonant contrasts equally well, but as they learn the phonological systems of their native language, this ability declines. Current explanations of this phenomenon agree that the decline in discrimination ability is linked to the formation of native-language phonemic categories. The goal of this study was to evaluate the role of input statistics in learning these categories: our hypothesis was that relative frequency is a determinant of the relative order in which categories are acquired. English-learning infants of two age groups (6.5 months and 8.5 months) were tested on their ability to discriminate non-native consonant contrasts using the Conditioned Head Turn Procedure. As predicted, older infants were worse in their performance on the more frequent coronal stop contrast than on the less frequent dorsal stop contrast. In contrast, 6.5-month-olds discriminated both contrasts equally well. An adult control group tested with an AX task also discriminated both contrasts equally. These results provide preliminary confirmation of the hypothesis that frequency plays an important role in tuning of phonological systems to properties of the native language. A simple attractor model suffices to account for these and previous results on loss of discrimination of non-native-language contrasts and suggests that the technique of measuring graded loss of multiple contrasts, in combination with observation of input frequencies, can offer a powerful method of assessing infants' phonological representations.
Discrimination Learning↗