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 721 records · Page 40Linked to original sources
PubMed2006

Neural representation of navigational relevance is rapidly induced and long lasting.

Successful navigation is facilitated by the presence of landmarks. Previous functional magnetic resonance imaging (fMRI) evidence indicated that the human parahippocampal gyrus automatically distinguishes between landmarks placed at navigationally relevant (decision points) and irrelevant locations (nondecision points). This storage of navigational relevance can provide a neural mechanism underlying successful navigation. However, an efficient wayfinding mechanism requires that important spatial information is learned quickly and maintained over time. The present study investigates whether the representation of navigational relevance is modulated by time and practice. Participants learned 2 film sequences through virtual mazes containing objects at decision and at nondecision points. One maze was shown one time, and the other maze was shown 3 times. Twenty-four hours after study, event-related fMRI data were acquired during recognition of the objects. The results showed that activity in the parahippocampal gyrus was increased for objects previously placed at decision points as compared with objects placed at nondecision points. The decision point effect was not modulated by the number of exposures to the mazes and independent of explicit memory functions. These findings suggest a persistent representation of navigationally relevant information, which is stable after only one exposure to an environment. These rapidly induced and long-lasting changes in object representation provide a basis for successful wayfinding.

Adolescent↗
PubMed1997

Text representations as reflected in patterns of cognitive distance.

Subjects performed a task that involved learning a text and then indicating, for each of a series of pairs of words, whether they belonged to the same sentence of that text. When the principle of argument overlap or causal coherence was the sole integrative factor in a text, negative response times and error rates reflected distances in the network representation predicted by the principle. When the two principles predicted conflicting representations for a text, response times and error rates reflected the predictions of the principle of argument overlap. The results were interpreted as suggesting that network connections predicted by the two principles are both present in the memory representations of texts and that causal connections are recorded at a more abstract level or representation than are argument overlap connections.

Adult↗
PubMed1986

Do semantic memory deficiencies underlie learning disabled readers' encoding processes?

This study investigates the extent to which learning disabled readers' atypical encoding relates to their deficiencies in semantic memory. Inferences related to ability group performance were based on the assumption that encoding involves the matching of incoming information against a featural representation of that information in semantic memory. To this end, learning disabled and nondisabled readers in two age groups were compared on dichotic listening recall tasks that included orienting and nonorienting instructions. Orienting instructions directed children's attention toward semantic, phonemic, or structural word features. Dependent measures were lateralization, free recall, retrieval organization, and selective attention. The efficiency of allocating attentional resources was inferred from correlations between central and incidental recall. Primary results included the following: Disabled and nondisabled readers' ear asymmetry differences were dependent upon age, orienting instructions, and type of word list; disabled readers' recall and organization scores were lower than skilled readers'; however, both ability groups benefitted from orienting instructions compared to nonorienting instructions; during orienting instructions, disabled readers were less able than skilled readers to divide their attention between target and nontarget word features, especially during interhemispheric processing conditions; and the relative efficiency of allocating attentional resources differed qualitatively between the two ability groups. The results suggest that ability group variations reflect the structure of the memory trace in interaction with ear presentation and encoding processes. It is inferred that disabled readers' inferior memory traces reflect the quantity and internal coherence of information stored in semantic memory as well as the means by which such information is accessed.

Attention↗
PubMed2002

Neural basis of mental scanning of a topographic representation built from a text.

Humans have the ability to build and to inspect an internal visual image of an environment built from a verbal description. We used positron emission tomography (PET) to investigate the brain areas engaged in the mental scanning of a map that subjects built from the reading of a descriptive text. This task engaged a parieto-frontal network known to deal with spatial representations. Additional activations were evidenced in the angular gyrus and in Broca's and Wernicke's areas. In order to examine the neural impact of the learning modality, these PET results were compared to those obtained in another group of six subjects who performed a similar mental scanning task on a topographic representation built from visual inspection of a map. Both scanning tasks engaged the parieto-frontal network. However, the bilateral activation of the angular gyrus as well as the involvement of language areas appeared specific to the mental scanning of the topographic representation built from textual information. On the other hand, the right medial temporal lobe was activated only when a map had been visually learned. These results suggest that although both tasks involved visuo-spatial internal representation, a trace of the learning modality remained present in the brain.

Adult↗
PubMed1999

Human simulations of vocabulary learning.

The work reported here experimentally investigates a striking generalization about vocabulary acquisition: Noun learning is superior to verb learning in the earliest moments of child language development. The dominant explanation of this phenomenon in the literature invokes differing conceptual requirements for items in these lexical categories: Verbs are cognitively more complex than nouns and so their acquisition must await certain mental developments in the infant. In the present work, we investigate an alternative hypothesis; namely, that it is the information requirements of verb learning, not the conceptual requirements, that crucially determine the acquisition order. Efficient verb learning requires access to structural features of the exposure language and thus cannot take place until a scaffolding of noun knowledge enables the acquisition of clause-level syntax. More generally, we experimentally investigate the hypothesis that vocabulary acquisition takes place via an incremental constraint-satisfaction procedure that bootstraps itself into successively more sophisticated linguistic representations which, in turn, enable new kinds of vocabulary learning. If the experimental subjects were young children, it would be difficult to distinguish between this information-centered hypothesis and the conceptual change hypothesis. Therefore the experimental "learners" are adults. The items to be "acquired" in the experiments were the 24 most frequent nouns and 24 most frequent verbs from a sample of maternal speech to 18-24-month-old infants. The various experiments ask about the kinds of information that will support identification of these words as they occur in mother-to-child discourse. Both the proportion correctly identified and the type of word that is identifiable changes significantly as a function of information type. We discuss these results as consistent with the incremental construction of a highly lexicalized grammar by cognitively and pragmatically sophisticated human infants, but inconsistent with a procedure in which lexical acquisition is independent of and antecedent to syntax acquisition.

Adult↗
PubMed2026

Leveraging Interradiomic Feature Relationships for Enhanced Prediction of Distant Metastasis and Characterization of Heterogeneity in Head and Neck Cancer.

PURPOSE: Distant metastasis remains a major cause of treatment failure in head and neck (HN) cancer, highlighting the need for more accurate early risk stratification. This study developed and validated a deep radiomics framework to characterize tumor heterogeneity from pretreatment computed tomography (CT) images and improve prediction of distant metastasis-free survival (DMFS). METHODS AND MATERIALS: This multicenter study included 3421 patients with HN cancer from 4 cohorts across 12 institutions. Radiomics features were extracted from primary tumors and transformed into OmicsMaps, a structured representation that spatially organizes interfeature relationships to facilitate learning of complex prognostic patterns. A convolutional neural network was trained to derive prognostic signatures, which were integrated with key clinical variables to construct an OmicsMap-clinical fusion model for patient risk stratification. Model performance was assessed using the concordance index (C-index) and time-dependent area under the receiver operating characteristic curve (AUC) in the CT Images from Large Head and Neck Cohort (RADCURE), HEAD-NECK-RADIOMICS-HN1 (HN1), and Head-Neck-Positron Emission Tomography-Computed Tomography (HN-PET-CT) cohorts. Radiogenomic analyses using RNA-seq data were conducted in the Cancer Genome Atlas Head-Neck Squamous Cell Carcinoma (TCGA-HNSC) cohort to investigate biological characteristics associated with the imaging-defined risk groups. RESULTS: The OmicsMap achieved C-index values of 0.742, 0.768, and 0.671 in the RADCURE, HN1, and HN-PET-CT cohorts, outperforming the conventional radiomics approach by 5.40%-6.37%. Incorporating clinical variables further improved generalizability, yielding a C-index of 0.864 (HN1) and 0.730 (HN-PET-CT), with time-dependent AUC of 0.727-0.895. The fusion model consistently stratified patients into distinct high- and low-risk groups for both DMFS and overall survival across cohorts (P <.01). Radiogenomic analyses revealed enrichment of immune-related pathways in the low-risk group, whereas the high-risk group exhibited a more aggressive phenotype enriched for proliferation, hypoxia, and epithelial-mesenchymal transition pathways, along with a fibrosis-prone tumor microenvironment characterized by extracellular matrix remodeling. CONCLUSIONS: Modeling interradiomic feature relationships using the OmicsMap representation substantially improves CT-based prediction of DMFS and characterization of tumor heterogeneity in HN cancer, supporting precision risk stratification in clinical oncology.

Journal Article↗
PubMed1995

Representation and performance: reproduction of form and quality of movement in dance by eight- and 11-year-old novices.

This study examined the effects of age on the cognitive representations in modelling. A total of 48 eight- and 11-year-olds were asked to watch a dancer performing a ballet sequence of three steps in order to reproduce it. They were able to see the film as many times as they wanted. All subjects were novices to dance. Both quantitative (number of observations required to learn the sequence, number of steps executed) and qualitative (form and quality of the performance) factors were assessed. The results indicated an effect of age on the number of requested observations and on the number of subjects able to perform the entire movement series. The mean form scores did not differ significantly between the two groups, but the mean quality score was higher for the 11-year-olds. We also noted a great heterogeneity in the performances. The results are discussed in terms of the role of cognitive and motor factors in observational learning by age and of the possibilities that cognitive representations offer for coding the different properties of movement.

Age Factors↗
PubMed2004

Experimental extinction in Pavlovian conditioning: behavioural and neuroscience perspectives.

This paper reviews the behavioural and neuroscience literatures on extinction in Pavlovian conditioning with a view towards finding possible points of contact between these two often independent lines of investigation. Recent discoveries at the behavioural level indicate (1) that conditioned stimulus (CS)-unconditioned stimulus (US) associations specific in their sensory content are fully preserved during extinction, (2) that inhibitory stimulus-response associations appear to be learned during extinction, (3) that extinction is influenced by the level of activation of the US representation during nonreinforced trials, (4) that decreases in attention can influence conditioned performance during extinction, and (5) that contexts acquire an ability to modulate learning during both conditioning and extinction. Recent discoveries at the neural systems level suggest (1) that the hippocampus is important in context-specific learning during extinction, (2) that the prefrontal cortex is possibly important in long-term memory for extinction, (3) that the basolateral amygdala may be important in sustaining attention to a CS during extinction, (4) that NMDA receptors are important either in neural plasticity during extinction or by affecting the value of the US representation during extinction, and (5) that the GABAergic system may partially mediate inhibitory learning during extinction. It is concluded that both of these levels of analysis can benefit the other in the pursuit of a more comprehensive understanding of extinction.

Amygdala↗
PubMed2005

Application of latent semantic analysis to protein remote homology detection.

MOTIVATION: Remote homology detection between protein sequences is a central problem in computational biology. The discriminative method such as the support vector machine (SVM) is one of the most effective methods. Many of the SVM-based methods focus on finding useful representations of protein sequence, using either explicit feature vector representations or kernel functions. Such representations may suffer from the peaking phenomenon in many machine-learning methods because the features are usually very large and noise data may be introduced. Based on these observations, this research focuses on feature extraction and efficient representation of protein vectors for SVM protein classification. RESULTS: In this study, a latent semantic analysis (LSA) model, which is an efficient feature extraction technique from natural language processing, has been introduced in protein remote homology detection. Several basic building blocks of protein sequences have been investigated as the 'words' of 'protein sequence language', including N-grams, patterns and motifs. Each protein sequence is taken as a 'document' that is composed of bags-of-word. The word-document matrix is constructed first. The LSA is performed on the matrix to produce the latent semantic representation vectors of protein sequences, leading to noise-removal and smart description of protein sequences. The latent semantic representation vectors are then evaluated by SVM. The method is tested on the SCOP 1.53 database. The results show that the LSA model significantly improves the performance of remote homology detection in comparison with the basic formalisms. Furthermore, the performance of this method is comparable with that of the complex kernel methods such as SVM-LA and better than that of other sequence-based methods such as PSI-BLAST and SVM-pairwise.

Algorithms↗
PubMed2005

Partially distributed representations of objects and faces in ventral temporal cortex.

Object and face representations in ventral temporal (VT) cortex were investigated by combining object confusability data from a computational model of object classification with neural response confusability data from a functional neuroimaging experiment. A pattern-based classification algorithm learned to categorize individual brain maps according to the object category being viewed by the subject. An identical algorithm learned to classify an image-based, view-dependent representation of the stimuli. High correlations were found between the confusability of object categories and the confusability of brain activity maps. This occurred even with the inclusion of multiple views of objects, and when the object classification model was tested with high spatial frequency "line drawings" of the stimuli. Consistent with a distributed representation of objects in VT cortex, the data indicate that object categories with shared image-based attributes have shared neural structure.

Algorithms↗
PubMed2005

Remapping auditory-motor representations in voice production.

Evidence regarding visually guided limb movements suggests that the motor system learns and maintains neural maps between motor commands and sensory feedback. Such systems are hypothesized to be used in a feed-forward control strategy that permits precision and stability without the delays of direct feedback control. Human vocalizations involve precise control over vocal and respiratory muscles. However, little is known about the sensorimotor representations underlying speech production. Here, we manipulated the heard fundamental frequency of the voice during speech to demonstrate learning of auditory-motor maps. Mandarin speakers repeatedly produced words with specific pitch patterns (tone categories). On each successive utterance, the frequency of their auditory feedback was increased by 1/100 of a semitone until they heard their feedback one full semitone above their true pitch. Subjects automatically compensated for these changes by lowering their vocal pitch. When feedback was unexpectedly returned to normal, speakers significantly increased the pitch of their productions beyond their initial baseline frequency. This adaptation was found to generalize to the production of another tone category. However, results indicate that a more robust adaptation was produced for the tone that was spoken during feedback alteration. The immediate aftereffects suggest a global remapping of the auditory-motor relationship after an extremely brief training period. However, this learning does not represent a complete transformation of the mapping; rather, it is in part target dependent.

Acoustic Stimulation↗
PubMed2001

Spatial representations of virtual mazes: the role of visual fidelity and individual differences.

Twenty-four people learned three versions of a room-sized maze: a wire-frame desktop virtual environment (VE), a normal surface-rendered desktop VE, and a real-world maze. Differences among the mental representations formed from each environment were measured with pointing and distance estimation tasks in a real-world version of each maze. People were more accurate at pointing after having learned the real and wire-frame VE maze than the surface-rendered VE maze; however, this effect was small compared with the effect of individual differences. Differences in gender, spatial ability, and prior computer experience were significantly related to the ability to acquire spatial information from the desktop VE. There was a high correlation between spatial knowledge when it was measured in the VE and spatial knowledge measured in the real world. Actual or potential applications include the design of effective VE training systems.

Adolescent↗
PubMed2005

Plasticity and reorganization of the uninjured brain.

Brain capacity is dependent not so much on the number of neurons but on the number of synaptic connections with functional connections that develop over a lifetime of genetic programming and life experiences. In the uninjured human brain, cortical reorganization that occurs in response to learning and experience is referred to as brain plasticity. Motor learning and complex environments result in a greater number of synapses and an increase in dendritic branching, whereas repetitive movements alone, in the absence of motor learning, do not. Learning and experience lead to an expansion of cortical representation, while failure to maintain training results in a contraction of cortical representation. In animals, loss of sensory peripheral afferent input results in an expansion of the forelimb representation of the intact adjacent cortex. Prolonged periods of peripheral nerve stimulation in both animals and humans can lead to reorganization of related sensorimotor cortical maps.

Animals↗
PubMed1998

A shared system for learning serial and temporal structure of sensori-motor sequences? Evidence from simulation and human experiments.

This research investigates the influences of temporal structure on the representation of serial order. Experiments are performed in a neural network model of sequence learning and in human subjects. In the sequence learning model, a recurrent network of leaky integrator neurons encodes a succession of internal states that become associated, by reinforcement learning, with the correct sequential responses. First, the model is shown to learn a simple temporal discrimination task. The model is then exposed to two novel serial reaction time (SRT) experiments. In the standard SRT task (M.J. Nissen, P. Bullemer, Attentional requirements of learning: evidence from performance measures, Cogn. Psychol. 19 (1987) 1-32 [16]), reaction times for stimuli presented in a repeating sequence are reduced with respect to those for random stimuli, providing a measure of sequence learning. The novelty of the current experiments is that imbedded in the serial order of the sequences, there is a temporal structure of delays. The model is sensitive to both the serial structure and the temporal structure of the sequences. This observation is then confirmed in human subjects. These results demonstrate how a novel recurrent architecture encodes the interaction of temporal and serial structure and provide insight into related aspects of human sensori-motor sequence learning.

Analysis of Variance↗
PubMed2006

Remembering the orientation of newly learned characters depends on the associated writing knowledge: a comparison between handwriting and typing.

Recent data support the idea that movements play a crucial role in letter representation and suggest that handwriting knowledge contributes to visual recognition of letters. If so, using different motor activities while subjects are learning to write should affect their subsequent recognition performances. In order to test this hypothesis, we trained adult participants to write new characters either by copying them or by typing them on a keyboard. After three weeks of training we ran a series of tests requiring visual processing of the characters' orientation. Tests were ran immediately, one week after, and three weeks after the end of the training period. Results showed that when the characters had been learned by typing, they were more frequently confused with their mirror images than when they had been written by hand. This handwriting advantage did not appear immediately, but mostly three weeks after the end of the training. Our results therefore suggest that the stability of the characters' representation in memory depends on the nature of the motor activity produced during learning.

Adult↗
PubMed2005

Neural mechanisms of autonomic, affective, and cognitive integration.

Influential theoretical models propose a central role for afferent information from the body in the expression of emotional feeling states. Feedback representations of changing states of bodily arousal influence learning and facilitate concurrent and prospective decision-making. Functional neuroimaging studies have increased understanding of brain mechanisms that generate changes in autonomic arousal during behavior and those which respond to internal feedback signals to influence subjective feeling states. In particular, anterior cingulate cortex is implicated in generating autonomic changes, while insula and orbitofrontal cortices may be specialized in mapping visceral responses. Independently, ventromedial prefrontal cortex is recognized to support processes of internal (self-) reference that predominate in states of rest and disengagement and which putatively serve as a benchmark for dynamic interactions with the environment. Lesion data further highlight the integrated role of these cortical regions in autonomic and motivational control. In computational models of control, forward (efference copies) and inverse models are proposed to enable prediction and correction of action and, by extension, the interpretation of the behavior of others. It is hypothesized that the neural substrate for these processes during motivational and affective behavior lies within the interactions of anterior cingulate, insula, and orbitofrontal cortices. Generation of visceral autonomic correlates of control reinforce experiential engagement in simulatory models and underpin concepts such as somatic markers to bridge the dualistic divide.

Afferent Pathways↗
PubMed2005

Classification of structural images via high-dimensional image warping, robust feature extraction, and SVM.

This paper presents a method for classification of medical images, using machine learning and deformation-based morphometry. A morphological representation of the anatomy of interest is first obtained using high-dimensional template warping, from which regions that display strong correlations between morphological measurements and the classification (clinical) variable are extracted using a watershed segmentation, taking into account the regional smoothness of the correlation map which is estimated by a cross-validation strategy in order to achieve robustness to outliers. A Support Vector Machine-Recursive Feature Elimination (SVM-RFE) technique is then used to rank computed features from the extracted regions, according to their effect on the leave-one-out error bound. Finally, SVM classification is applied using the best set of features, and it is tested using leave-one-out. The results from a group of 61 brain images of female normal controls and schizophrenia patients demonstrate not only high classification accuracy (91.8%) and steep ROC curves, but also exceptional stability with respect to the number of selected features and the SVM kernel size.

Algorithms↗
PubMed1986

Linear function neurons: structure and training.

Three different representations for a thresholded linear equation are developed. For binary input they are shown to be representationally equivalent though their training characteristics differ. A training algorithm for linear equations is discussed. The similarities between its simplest mathematical representation (perceptron training), a formal model of animal learning (Rescorla-Wagner learning), and one mechanism of neural learning (Aplysia gill withdrawal) are pointed out. For d input features, perceptron training is shown to have a lower bound of 2d and an upper bound of dd adjusts. It is possible that the true upper bound is 4d, though this has not been proved. Average performance is shown to have a lower bound of 1.4d. Learning time is shown to increase linearly with the number of irrelevant or replicated features. The (X of N) function (a subset of linearly separable functions containing OR and AND) is shown to be learnable in d3 time. A method of utilizing conditional probability to accelerate learning is proposed. This reduces the observed growth rate from 4d to the theoretical minimum (for unmodified version) of 2d. A different version reduces the growth rate to about 1.7d. The linear effect of irrelevant features can also be eliminated. Whether such an approach can be made probably convergent is not known.

Animals↗