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 991 records · Page 55Linked to original sources
PubMedDate not supplied

[Role of different projection areas of the motor cortex in reorganization of the innate head-forelimb coordination in dogs].

Dogs were trained to perform the forelimb tonic flexion in order to lift a cup with meat from a bottom of the foodwell and hold it during eating with the head bent down to the cup. It is known that conditioning of the instrumental reaction is based on reorganization of the innate head-forelimb coordination into the opposite one. In untrained dogs, the forelimb flexion is accompanied by the anticipatory lifting of the head bent down to the foodwell. The following lowering of the head leads to an extension of the flexed forelimb. Tonic forelimb flexion is possible if the head is in the up position. Simultaneous holding of the flexed forelimb and lowered head providing food reinforcement is achieved only by learning. It was shown earlier that the lesion of the motor cortex contralateral to the "working" forelimb led to a prolonged disturbance of the elaborated coordination and reappearance of the innate coordination. In the present work we studied the influence of local lesions of the projection areas in the motor cortex, such as a "working" forelimb area, bilateral representation of the neck, and the medial part of the motor cortex, on the learned instrumental feeding reaction. It was found that only the lesion of the forelimb but not neck projection led to a disturbance of the learned head-forelimb movement coordination.

Animals↗
PubMedDate not supplied

Neural representations of location outside the hippocampus.

Place cells of the rat hippocampus are a dominant model system for understanding the role of the hippocampus in learning and memory at the level of single-unit and neural ensemble responses. A complete understanding of the information processing and computations performed by the hippocampus requires detailed knowledge about the properties of the representations that are present in hippocampal afferents and efferents in order to decipher the transformations that occur to these representations in the hippocampal circuitry. Neural recordings in behaving rats have revealed a number of brain areas that contain place-related firing properties in the parahippocampal regions and in other brain regions that are thought to interact with the hippocampus in certain behavioral tasks. Although investigators have just begun to scratch the surface in terms of understanding these properties, differences in the precise nature of the spatial firing between the hippocampus and these other regions promise to reveal important clues regarding the exact role of the hippocampus in learning and memory and the nature of its interactions with other brain systems to support adaptive behavior.

Animals↗
PubMed2004

Intelligent automated control of life support systems using proportional representations.

Effective automatic control of Advanced Life Support Systems (ALSS) is a crucial component of space exploration. An ALSS is a coupled dynamical system which can be extremely sensitive and difficult to predict. As a result, such systems can be difficult to control using deliberative and deterministic methods. We investigate the performance of two machine learning algorithms, a genetic algorithm (GA) and a stochastic hill-climber (SH), on the problem of learning how to control an ALSS, and compare the impact of two different types of problem representations on the performance of both algorithms. We perform experiments on three ALSS optimization problems using five strategies with multiple variations of a proportional representation for a total of 120 experiments. Results indicate that although a proportional representation can effectively boost GA performance, it does not necessarily have the same effect on other algorithms such as SH. Results also support previous conclusions that multivector control strategies are an effective method for control of coupled dynamical systems.

Algorithms↗
PubMed2007

Lexical restructuring in the absence of literacy.

Vocabulary growth was suggested to prompt the implementation of increasingly finer-grained lexical representations of spoken words in children (e.g., [Metsala, J. L., & Walley, A. C. (1998). Spoken vocabulary growth and the segmental restructuring of lexical representations: precursors to phonemic awareness and early reading ability. In J. L. Metsala & L. C. Ehri (Eds.), Word recognition in beginning literacy (pp. 89-120). Hillsdale, NJ: Erlbaum.]). Although literacy was not explicitly mentioned in this lexical restructuring hypothesis, the process of learning to read and spell might also have a significant impact on the specification of lexical representations (e.g., [Carroll, J. M., & Snowling, M. J. (2001). The effects of global similarity between stimuli on children's judgments of rime and alliteration. Applied Psycholinguistics, 22, 327-342.]; [Goswami, U. (2000). Phonological representations, reading development and dyslexia: Towards a cross-linguistic theoretical framework. Dyslexia, 6, 133-151.]). This is what we checked in the present study. We manipulated word frequency and neighborhood density in a gating task (Experiment 1) and a word-identification-in-noise task (Experiment 2) presented to Portuguese literate and illiterate adults. Ex-illiterates were also tested in Experiment 2 in order to disentangle the effects of vocabulary size and literacy. There was an interaction between word frequency and neighborhood density, which was similar in the three groups. These did not differ even for the words that are supposed to undergo lexical restructuring the latest (low frequency words from sparse neighborhoods). Thus, segmental lexical representations seem to develop independently of literacy. While segmental restructuring is not affected by literacy, it constrains the development of phoneme awareness as shown by the fact that, in Experiment 3, neighborhood density modulated the phoneme deletion performance of both illiterates and ex-illiterates.

Adolescent↗
PubMed2025

Leveraging protein language models for cross-variant CRISPR/Cas9 sgRNA activity prediction.

MOTIVATION: Accurate prediction of single-guide RNA (sgRNA) activity is crucial for optimizing the CRISPR/Cas9 gene-editing system, as it directly influences the efficiency and accuracy of genome modifications. However, existing prediction methods mainly rely on large-scale experimental data of a single Cas9 variant to construct Cas9 protein (variants)-specific sgRNA activity prediction models, which limits their generalization ability and prediction performance across different Cas9 protein (variants), as well as their scalability to the continuously discovered new variants. RESULTS: In this study, we proposed PLM-CRISPR, a novel deep learning-based model that leverages protein language models to capture Cas9 protein (variants) representations for cross-variant sgRNA activity prediction. PLM-CRISPR uses tailored feature extraction modules for both sgRNA and protein sequences, incorporating a cross-variant training strategy and a dynamic feature fusion mechanism to effectively model their interactions. Extensive experiments demonstrate that PLM-CRISPR outperforms existing methods across datasets spanning seven Cas9 protein (variants) in three real-world scenarios, demonstrating its superior performance in handling data-scarce situations, including cases with few or no samples for novel variants. Comparative analyses with traditional machine learning and deep learning models further confirm the effectiveness of PLM-CRISPR. Additionally, motif analysis reveals that PLM-CRISPR accurately identifies high-activity sgRNA sequence patterns across diverse Cas9 protein (variants). Overall, PLM-CRISPR provides a robust, scalable, and generalizable solution for sgRNA activity prediction across diverse Cas9 protein (variants). AVAILABILITY AND IMPLEMENTATION: The source code can be obtained from https://github.com/CSUBioGroup/PLM-CRISPR.

CRISPR-Cas Systems↗
PubMedDate not supplied

Reflections on human Pavlovian decelerative heart-rate conditioning with negative tilt as US: alternative approaches.

The negative-tilt preparation that has been reported since the late seventies is a specific form of Pavlovian conditioning that is of scientific interest and has potential applications. In this paper I reflect on the usefulness, to the development of this preparation, of two approaches to Pavlovian conditioning. One approach is the older S-R learning, stimulus-substitution paradigm exemplified by learning texts of the sixties. The other is the modern, Tolman-like view, according to which the phenomenon of Pavlovian conditioning is "now described as the learning of relations among events so as to allow the organism to represent its environment." The three assumptions encapsulated by this approach are: (a) that only CS-US contingency relations are learned; (b) that teleological modes of explanations are adequate; (c) that the representational theory of knowledge is sound. Concerning Pavlovian conditioning in general, questions been raised in the literature for all three assumptions; they have not been adequately answered. Regarding the specific problem of developing the human Pavlovian heart-rate decelerative conditioning with negative tilt as the US, I suggest that the cognitive approach has been much less helpful than the older, S-R, stimulus-substitution paradigm. Nevertheless, other literature clearly indicates that the cognitive, S-S approach has generated considerable interest and research, especially in preparations like the conditioned emotional response (CER), which are CS-IR ones in the sense that the effects on the CR are assessed indirectly through measuring an indicator or instrumental response (IR).(ABSTRACT TRUNCATED AT 250 WORDS)

Animals↗
PubMed2002

View from the top: hierarchies and reverse hierarchies in the visual system.

We propose that explicit vision advances in reverse hierarchical direction, as shown for perceptual learning. Processing along the feedforward hierarchy of areas, leading to increasingly complex representations, is automatic and implicit, while conscious perception begins at the hierarchy's top, gradually returning downward as needed. Thus, our initial conscious percept--vision at a glance--matches a high-level, generalized, categorical scene interpretation, identifying "forest before trees." For later vision with scrutiny, reverse hierarchy routines focus attention to specific, active, low-level units, incorporating into conscious perception detailed information available there. Reverse Hierarchy Theory dissociates between early explicit perception and implicit low-level vision, explaining a variety of phenomena. Feature search "pop-out" is attributed to high areas, where large receptive fields underlie spread attention detecting categorical differences. Search for conjunctions or fine discriminations depends on reentry to low-level specific receptive fields using serial focused attention, consistent with recently reported primary visual cortex effects.

Humans↗
PubMed2004

Visual, haptic and crossmodal recognition of scenes.

Real-world scene perception can often involve more than one sensory modality. Here we investigated the visual, haptic and crossmodal recognition of scenes of familiar objects. In three experiments participants first learned a scene of objects arranged in random positions on a platform. After learning, the experimenter swapped the position of two objects in the scene and the task for the participant was to identify the two swapped objects. In experiment 1, we found a cost in scene recognition performance when there was a change in sensory modality and scene orientation between learning and test. The cost in crossmodal performance was not due to the participants verbally encoding the objects (experiment 2) or by differences between serial and parallel encoding of the objects during haptic and visual learning, respectively (experiment 3). Instead, our findings suggest that differences between visual and haptic representations of space may affect the recognition of scenes of objects across these modalities.

Adult↗
PubMed2002

Transfer of motor learning across arm configurations.

It has been suggested that the learning of new dynamics occurs in intrinsic coordinates. However, it has also been suggested that elements that encode hand velocity, and hence act in an extrinsic frame of reference, play a role in the acquisition of dynamics. To reconcile claims regarding the coordinate system involved in the representation of dynamics, we have used a procedure involving the transfer of force-field learning between two workspace locations. Subjects made point-to-point movements while holding a two-link manipulandum. Subjects were first trained to make movements in a single direction at the left of the workspace. They were then tested for transfer of learning at the right of the workspace. Two groups of subjects were defined. For the subjects in group j, movements at the left and right workspace locations were matched in terms of joint displacements. For the subjects in group h, movements in the two locations had the same hand displacements. Workspace locations were chosen such that for group j, the paths (for training and testing) that were identical in joint space were orthogonal in hand space. The subjects in group j showed good transfer between workspace locations, whereas the subjects in group h showed poor transfer. These results are in agreement with the idea that new dynamics are encoded in intrinsic coordinates and that this learning has a limited range of generalization across joint velocities.

Adult↗
PubMed2005

A comparison of algorithms for inference and learning in probabilistic graphical models.

Research into methods for reasoning under uncertainty is currently one of the most exciting areas of artificial intelligence, largely because it has recently become possible to record, store, and process large amounts of data. While impressive achievements have been made in pattern classification problems such as handwritten character recognition, face detection, speaker identification, and prediction of gene function, it is even more exciting that researchers are on the verge of introducing systems that can perform large-scale combinatorial analyses of data, decomposing the data into interacting components. For example, computational methods for automatic scene analysis are now emerging in the computer vision community. These methods decompose an input image into its constituent objects, lighting conditions, motion patterns, etc. Two of the main challenges are finding effective representations and models in specific applications and finding efficient algorithms for inference and learning in these models. In this paper, we advocate the use of graph-based probability models and their associated inference and learning algorithms. We review exact techniques and various approximate, computationally efficient techniques, including iterated conditional modes, the expectation maximization (EM) algorithm, Gibbs sampling, the mean field method, variational techniques, structured variational techniques and the sum-product algorithm ("loopy" belief propagation). We describe how each technique can be applied in a vision model of multiple, occluding objects and contrast the behaviors and performances of the techniques using a unifying cost function, free energy.

Algorithms↗
PubMed2002

Extending the ALCOVE model of category learning to featural stimulus domains.

The ALCOVE model of category learning, despite its considerable success in accounting for human performance across a wide range of empirical tasks, is limited by its reliance on spatial stimulus representations. Some stimulus domains are better suited to featural representation, characterizing stimuli in terms of the presence or absence of discrete features, rather than as points in a multidimensional space. We report on empirical data measuring human categorization performance across a featural stimulus domain and show that ALCOVE is unable to capture fundamental qualitative aspects of this performance. In response, a featural version of the ALCOVE model is developed, replacing the spatial stimulus representations that are usually generated by multidimensional scaling with featural representations generated by additive clustering. We demonstrate that this featural version of ALCOVE is able to capture human performance where the spatial model failed, explaining the difference in terms of the contrasting representational assumptions made by the two approaches. Finally, we discuss ways in which the ALCOVE categorization model might be extended further to use "hybrid" representational structures combining spatial and featural components.

Discrimination Learning↗
PubMed2004

The interactive effects of prior knowledge and text structure on memory for cognitive psychology texts.

BACKGROUND: Interest in the interactive effects of prior knowledge and text structure on learning from text is increasing but experimental manipulations of knowledge and structure variables often produce findings that do not help teachers to select expository texts for students. AIMS: We aimed to extend the ecological validity of previous findings by asking students with a high or low level of discipline-relevant knowledge to read texts characteristic of those they would normally encounter. A compensation effect was hypothesized, where high prior knowledge would compensate for a lack of text structure and text structure would compensate for a lack of prior knowledge. SAMPLES: One hundred and ninety-five undergraduate psychology students (144 Year 1 students and 51 Year 3 students) were allocated to a high knowledge (HiPK) or low knowledge (LoPK) group on the basis of their performance on a word association test. METHODS: Participants were randomly assigned to one of five text structure groups (compare/contrast, sequence, classification, enumeration, generalization) and asked to study two cognitive psychology passages before recalling the main points of the text immediately afterwards and after a delay of 2 weeks. The five text structures were placed on an 'organizational continuum' according to the degree of structure/organization in the 10 passages. RESULTS: A compensation effect did not emerge. Recall was high when texts were well structured and readers had prior knowledge, but recall was poor when texts were less structured, regardless of the level of prior knowledge. CONCLUSIONS: Readers benefit most from texts that challenge pre-existing mental representations.

Awareness↗
PubMed1998

What is the role of dopamine in reward: hedonic impact, reward learning, or incentive salience?

What roles do mesolimbic and neostriatal dopamine systems play in reward? Do they mediate the hedonic impact of rewarding stimuli? Do they mediate hedonic reward learning and associative prediction? Our review of the literature, together with results of a new study of residual reward capacity after dopamine depletion, indicates the answer to both questions is 'no'. Rather, dopamine systems may mediate the incentive salience of rewards, modulating their motivational value in a manner separable from hedonia and reward learning. In a study of the consequences of dopamine loss, rats were depleted of dopamine in the nucleus accumbens and neostriatum by up to 99% using 6-hydroxydopamine. In a series of experiments, we applied the 'taste reactivity' measure of affective reactions (gapes, etc.) to assess the capacity of dopamine-depleted rats for: 1) normal affect (hedonic and aversive reactions), 2) modulation of hedonic affect by associative learning (taste aversion conditioning), and 3) hedonic enhancement of affect by non-dopaminergic pharmacological manipulation of palatability (benzodiazepine administration). We found normal hedonic reaction patterns to sucrose vs. quinine, normal learning of new hedonic stimulus values (a change in palatability based on predictive relations), and normal pharmacological hedonic enhancement of palatability. We discuss these results in the context of hypotheses and data concerning the role of dopamine in reward. We review neurochemical, electrophysiological, and other behavioral evidence. We conclude that dopamine systems are not needed either to mediate the hedonic pleasure of reinforcers or to mediate predictive associations involved in hedonic reward learning. We conclude instead that dopamine may be more important to incentive salience attributions to the neural representations of reward-related stimuli. Incentive salience, we suggest, is a distinct component of motivation and reward. In other words, dopamine systems are necessary for 'wanting' incentives, but not for 'liking' them or for learning new 'likes' and 'dislikes'.

Animals↗
PubMed1996

Eliciting facial affect, motivation, and expectancies in transference: significant-other representations in social relations.

Recent research has demonstrated transference in social perception, defined in terms of memory and schema-triggered evaluation in relation to a new person (S. M. Andersen & A. B. Baum, 1994; S. M. Andersen & S. W. Cole, 1990; S. M. Andersen, N. S. Glassman, S. Chen, & S. W. Cole, 1995). The authors examined schema-triggered facial affect in transference, along with motivations and expectancies. In a nomothetic experimental design, participants encountered stimulus descriptors of a new target person that were derived either from their own idiographic descriptions of a positively toned or a negatively toned significant other or from a yoked control participant's descriptors. Equal numbers of positive and negative target descriptors were presented, regardless of the overall tone of the representation. The results verified the memory effect and schema-triggered evaluation in transference, on the basis of significant-other resemblance in the target person. Of importance, participants' nonverbal expression of facial affect when learning about the target person (i.e., at encoding) reflected the overall tone of their significant-other representation under the condition of significant-other resemblance, providing strong support for schema-triggered affect in transference, through the use of this unobtrusive, nonverbal measure. Parallel effects on interpersonal closeness motivation and expectancies for acceptance/rejection in transference also emerged.

Affect↗
PubMed1990

Implicit learning in patients with Alzheimer's disease.

We examined implicit memory using priming and procedural learning tasks in patients with probable Dementia-Alzheimer's Type (DAT) to examine whether priming and procedural processes could be dissociated and whether task specificity was a factor in DAT patient performance. Priming was tested using a word recognition paradigm (perceptual priming) and by repeated administrations of a fragmented objects test (long term priming). Procedural learning was tested using repeated and random sequences on a choice serial reaction time task and by repeated administration of a puzzle map of the United States. DAT patients were compared to hospitalized depressed patients, patients suffering from Progressive Supranuclear Palsy (PSP), and normal controls. We found that DAT patients demonstrated marginal but significant implicit learning on both procedural learning and perceptual priming tasks. DAT patients performed relatively better on the procedural learning task than a perceptual priming task compared to PSP patients, suggesting that priming of meaningful stimuli is subserved by cortical structures whereas procedural motor responses to simple serial visual stimulus patterns can be maintained by subcortical systems. Furthermore, our findings suggest that priming and procedural processes can be dissociated and that task specificity is a factor in interpreting the results of implicit learning paradigms in DAT patients. The implications of these results for models of knowledge representation and memory processes as well as the way they can serve as models for testing nootropic drug effects are discussed.

Aged↗
PubMed1997

Behavioral characterization of metrifonate-improved acquisition of spatial information in medial septum-lesioned rats.

We investigated the effects of acute oral pretraining treatment with an indirect acetylcholinesterase inhibitor, metrifonate, on water maze spatial navigation in medial septum-lesioned rats. We observed that metrifonate (30 mg/kg, orally) (1) does not alter the pattern of exploration of lesioned rats at the water maze pool or retrieval of spatial memory, (2) effectively reverses the acquisition defect, (3) enhances reversal learning, and (4) improves acquisition of water maze navigation by facilitating the encoding of the spatial representation of a specific environment. These results indicate that metrifonate does not improve escape performance to the hidden platform by modulating exploration strategy, but that metrifonate enhances the speed and accuracy of development and durability of spatial memory engrams, and facilitates learning capacity that depends on activity of the septo-hippocampal projection.

Acetylcholinesterase↗
PubMed2004

Knowledge acquisition in the fuzzy knowledge representation framework of a medical consultation system.

This paper describes the fuzzy knowledge representation framework of the medical computer consultation system MedFrame/CADIAG-IV as well as the specific knowledge acquisition techniques that have been developed to support the definition of knowledge concepts and inference rules. As in its predecessor system CADIAG-II, fuzzy medical knowledge bases are used to model the uncertainty and the vagueness of medical concepts and fuzzy logic reasoning mechanisms provide the basic inference processes. The elicitation and acquisition of medical knowledge from domain experts has often been described as the most difficult and time-consuming task in knowledge-based system development in medicine. It comes as no surprise that this is even more so when unfamiliar representations like fuzzy membership functions are to be acquired. From previous projects we have learned that a user-centered approach is mandatory in complex and ill-defined knowledge domains such as internal medicine. This paper describes the knowledge acquisition framework that has been developed in order to make easier and more accessible the three main tasks of: (a) defining medical concepts; (b) providing appropriate interpretations for patient data; and (c) constructing inferential knowledge in a fuzzy knowledge representation framework. Special emphasis is laid on the motivations for some system design and data modeling decisions. The theoretical framework has been implemented in a software package, the Knowledge Base Builder Toolkit. The conception and the design of this system reflect the need for a user-centered, intuitive, and easy-to-handle tool. First results gained from pilot studies have shown that our approach can be successfully implemented in the context of a complex fuzzy theoretical framework. As a result, this critical aspect of knowledge-based system development can be accomplished more easily.

Decision Making↗
PubMed2002

Early consolidation of instrumental learning requires protein synthesis in the nucleus accumbens.

It is widely held that long-term memories are established by consolidation of newly acquired information into stable neural representations, a process that requires protein synthesis and synaptic plasticity. Plasticity within the nucleus accumbens (NAc), a major component of the ventral striatum, is thought to mediate instrumental learning processes and many aspects of drug addiction. Here we show that the inhibition of protein synthesis within the NAc disrupts consolidation of an appetitive instrumental learning task (lever-pressing for food) in rats. Post-trial infusions of anisomycin immediately after the first several training sessions prevented consolidation, whereas infusions delayed by 2 or 4 hours had no effect. However, if the rats were allowed to learn the task, the behavior was not sensitive to disruption by intra-accumbens anisomycin. Control infusions into the medial NAc shell or the dorsolateral striatum did not impair learning; in fact, an enhancement was observed in the latter case. These results show that de novo protein synthesis within the NAc is necessary for the consolidation, but not reconsolidation, of appetitive instrumental memories.

Animals↗