Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Ensemble 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 253 records · Page 14Linked to original sources
PubMed2002

Genetically expressed cameleon in Drosophila melanogaster is used to visualize olfactory information in projection neurons.

Complex external stimuli such as odorants are believed to be internally represented in the brain by spatiotemporal activity patterns of extensive neuronal ensembles. These activity patterns can be recorded by optical imaging techniques. However, optical imaging with conventional fluorescence dyes usually does not allow for resolving the activity of biologically defined groups of neurons. Therefore, specifically targeting reporter molecules to neuron populations of common genetic identity is an important goal. We report the use of the genetically encoded calcium-sensitive fluorescence protein cameleon 2.1 in the Drosophila brain. We visualized odorant-evoked intracellular calcium concentration changes in selectively labeled olfactory projection neurons both postsynaptically in the antennal lobe, the primary olfactory neuropil, and presynaptically in the mushroom body calyx, a structure involved in olfactory learning and memory. As a technical achievement, we show that calcium imaging with a genetically encoded fluorescence probe is feasible in a brain in vivo. This will allow one to combine Drosophila's advanced genetic tools with the physiological analysis of brain function. Moreover, we report for the first time optical imaging recordings in synaptic regions of the Drosophila mushroom body calyx and antennal lobe. This provides an important step for the use of Drosophila as a model system in olfaction.

Animals↗
PubMed1996

Neurons, networks, and motor behavior.

The field of motor pattern generation and motor control has progressed markedly in the last decade. There has been a revolutionary shift in thinking from hard-wired circuits to multifunctional networks. Yet, it is clear that we still have a long way to go before we understand how very large ensembles of neurons produce behaviors. The systems where we have made the most headway are those that have an orderly topography, such as the superior colliculus (Sparks) or motor cortex (Georgopoulos). However, even in these systems, although we understand how to interpret the combined activity of the neuronal population, it is not clear how this population activity is translated into a motor command. Similarly, the directional behavior produced in cockroach (Ritzmann) and fish escape (Eaton) systems can be predicted based on the activity of neurons, but the cellular mechanisms producing the turning responses in cockroaches and teleost fish are not completely understood. Undoubtedly, computational approaches, including new mathematical formalisms and computer simulations, will play a role in elucidating how very large ensembles of neurons produce their coordinated output. For now, the systems where motor pattern generation is best understood at the cellular level are those with small numbers of neurons (such as invertebrate circuits) or small numbers of cell types, such as lamprey and tadpole spinal circuits. These systems are thus valuable for pointing to potential mechanisms used in larger systems. (Note that I avoid using the term "simple" systems to describe invertebrates because it is quite clear that these systems are anything but simple.) However, "interphyletic awareness," as it was referred to at this conference, is not important just for what it can tell us about how mammals work. It is also important to learn of alternative ways in which organisms solve similar problems. This may prove to be particularly important for the future of robotics. Already, robots have been designed based on insights gained from studying insect visual (Strausfeld) and motor (Ritzmann) systems. Robotics engineers have also independently converged on some of the same mechanisms used by biological systems (MacPherson). There is clearly a need for better understanding of higher control of pattern-generating circuits. This is not limited to how motor patterns are initiated, but also includes how they are altered on a moment to moment basis to suit the needs of the animal. The next revolution in the field is likely to come from a paradigm shift regarding such control of motor circuits, similar to the shift that has already occurred in our understanding of the pattern-generating circuits themselves. Such flexibility of control is the basis for decision making in the nervous system and the very essence of what animals must do throughout their daily lives. I look forward to the next conference in 2005 to see how far we've progressed in these pursuits.

Animals↗
PubMed2006

Predicting protein subcellular locations using hierarchical ensemble of Bayesian classifiers based on Markov chains.

BACKGROUND: The subcellular location of a protein is closely related to its function. It would be worthwhile to develop a method to predict the subcellular location for a given protein when only the amino acid sequence of the protein is known. Although many efforts have been made to predict subcellular location from sequence information only, there is the need for further research to improve the accuracy of prediction. RESULTS: A novel method called HensBC is introduced to predict protein subcellular location. HensBC is a recursive algorithm which constructs a hierarchical ensemble of classifiers. The classifiers used are Bayesian classifiers based on Markov chain models. We tested our method on six various datasets; among them are Gram-negative bacteria dataset, data for discriminating outer membrane proteins and apoptosis proteins dataset. We observed that our method can predict the subcellular location with high accuracy. Another advantage of the proposed method is that it can improve the accuracy of the prediction of some classes with few sequences in training and is therefore useful for datasets with imbalanced distribution of classes. CONCLUSION: This study introduces an algorithm which uses only the primary sequence of a protein to predict its subcellular location. The proposed recursive scheme represents an interesting methodology for learning and combining classifiers. The method is computationally efficient and competitive with the previously reported approaches in terms of prediction accuracies as empirical results indicate. The code for the software is available upon request.

Amino Acid Sequence↗
PubMed1999

A tracing evoked potential estimator.

The paper presents an adaptive Gaussian radial basis function neural network (RBFNN) for rapid estimation of evoked potential (EP). Usually, a recorded EP is severely contaminated by background ongoing activities of the brain. Many approaches have been reported to enhance the signal-to-noise ratio (SNR) of the recorded signal. However, non-linear methods are seldom explored due to their complexity and the fact that the non-linear characteristics of the signal are generally hard to determine. An RBFNN possesses built-in non-linear activation functions that enable the neural network to learn any function mapping. An RBFNN was carefully designed to model the EP signal. It has the advantage of being linear-in-parameter, thus a conventional adaptive method can efficiently estimate its parameters. The proposed algorithm is simple so that its convergence behaviour and performance in signal-to-noise ratio (SNR) improvement can be mathematically derived. A series of experiments carried out on simulated and human test responses confirmed the superior performance of the method. In a simulation experiment, an RBFNN having 15 hidden nodes was trained to approximate human visual EP (VEP). For detecting human brain stem auditory EP (BAEP), the approach (40 hidden nodes and convergence rate = 0.005) speeded up the estimation remarkably by using only 80 ensembles to achieve a result comparable to that obtained by averaging 1000 ensembles.

Electroencephalography↗
PubMed2003

Prospective and retrospective memory coding in the hippocampus.

The effect of memory on hippocampal neuronal activity was assessed as rats performed a spatial task that was impaired by fornix lesions. The influences of current location, recently entered places, and places about to be entered were compared. Three new findings emerged. (1) Current, retrospective, and prospective coding were common and recorded simultaneously in neural ensembles. (2) The origin of journeys influenced firing even when rats made detours, showing that recent memory could modulate neuronal activity more than spatial trajectory. (3) Diminished retrospective coding and, more markedly, reduced prospective coding in error trials suggested that the neuronal signal was important for task performance. The population of hippocampal neurons thus encoded information about the recent past, the present, and the imminent future, consistent with a neuronal mechanism for episodic memory.

Animals↗
PubMed1996

A model of spatial map formation in the hippocampus of the rat.

Using experimental facts about long-term potentiation (LTP) and hippocampal place cells, we model how a spatial map of the environment can be created in the rat hippocampus. Sequential firing of place cells during exploration induces, in the model, a pattern of LTP between place cells that shifts the location coded by their ensemble activity away from the actual location of the animal. These shifts provide a navigational map that, in a simulation of the Morris maze, can guide the animal toward its goal. The model demonstrates how behaviorally generated modifications of synaptic strengths can be read out to affect subsequent behavior. Our results also suggest a way that navigational maps can be constructed from experimental recordings of hippocampal place cells.

Action Potentials↗
PubMed2005

Bio-basis function neural network for prediction of protease cleavage sites in proteins.

The prediction of protease cleavage sites in proteins is critical to effective drug design. One of the important issues in constructing an accurate and efficient predictor is how to present nonnumerical amino acids to a model effectively. As this issue has not yet been paid full attention and is closely related to model efficiency and accuracy, we present a novel neural learning algorithm aimed at improving the prediction accuracy and reducing the time involved in training. The algorithm is developed based on the conventional radial basis function neural networks (RBFNNs) and is referred to as a bio-basis function neural network (BBFNN). The basic principle is to replace the radial basis function used in RBFNNs by a novel bio-basis function. Each bio-basis is a feature dimension in a numerical feature space, to which a nonnumerical sequence space is mapped for analysis. The bio-basis function is designed using an amino acid mutation matrix verified in biology. Thus, the biological content in protein sequences can be maximally utilized for accurate modeling. Mutual information (MI) is used to select the most informative bio-bases and an ensemble method is used to enhance a decision-making process, hence, improving the prediction accuracy further. The algorithm has been successfully verified in two case studies, namely the prediction of Human Immunodeficiency Virus (HIV) protease cleavage sites and trypsin cleavage sites in proteins.

Artificial Intelligence↗
PubMed2006

Fast rate coding in hippocampal CA3 cell ensembles.

Environments with overlapping features are represented by distinct patterns of activity in the hippocampus, enabling information to be stored and retrieved with minimal interference. This orthogonalization of correlated inputs is thought to take place within the hippocampus itself. However, the orthogonalization process has been shown to take days to develop in CA1. This prolonged time course is in striking contrast to the fast encoding of behavioral memory by the hippocampus. To explore this apparent paradox, we asked whether orthogonalization depended on the type of remapping exhibited by the hippocampal network. We have previously distinguished two types of remapping, global remapping, which results in the activation of different assemblies of place fields, and rate remapping, which encodes differences between cue constellations by substantial changes in firing rate without a change in the place code. Global remapping has previously been shown to be expressed immediately at novel locations. Here we asked if rate remapping follows a slower time course. Ensemble activity was recorded simultaneously from CA3 and CA1 in rats exposed to two similar, novel environments. It was found that rate changes in response to novel sensory cue configurations can form immediately, just as during global remapping, in particular in CA3. The fast encoding of both spatial and nonspatial information in CA3 is consistent with a role for the autoassociative CA3 circuitry in the acquisition and expression of episodic memories.

Action Potentials↗
PubMed1994

Zaidel's model of interhemispheric dynamics: empirical tests, a critical appraisal, and a proposed revision.

Zaidel has proposed a model of intra- and interhemispheric dynamics which attempts to explain field and hand effects in visual reaction time experiments. The model includes prototypes of two extreme cases: a "callosal relay" pattern, in which interhemispheric transfer is inferred, and a "direct access" pattern, in which each hemisphere is inferred to execute the task independently. Five visual reaction time experiments were carried out to test six predictions which Zaidel has made within the framework of his model. Some of these predictions were supported but others were rejected. These findings, in conjunction with findings of other researchers, were critically analyzed in view of determining the extent to which Zaidel's model can be revised in order to fit the ensemble of relevant findings. It was concluded that the model can indeed be revised but that this requires a redefinition of conditions indicative of presence or absence of interhemispheric relay. A new model is proposed assuming exclusively excitatory callosal relay and focussing on magnitude and valence of field effects obtained at each hand, considered in their interrelation.

Adult↗
PubMed2006

Hippocampal sharp waves and reactivation during awake states depend on repeated sequential experience.

Hippocampal firing patterns during behavior are reactivated during rest and subsequent slow-wave sleep. These reactivations occur during transient local field potential (LFP) events, termed sharp waves. Theories of hippocampal processing suggest that sharp waves arise from strengthened plasticity, and that the strengthened plasticity depends on repeated cofiring of pyramidal cells. We tested these predictions by recording neural ensembles and LFPs from rats running tasks requiring different levels of behavioral repetition. The number of sharp waves emitted increased during sessions with more regular behaviors. Reactivation became more similar to behavioral firing patterns across the session. This enhanced reactivation also depended on the regularity of the behavior. Additional studies in CA3 and CA1 found that the number of sharp waves emitted also increased in CA3 recordings as well as CA1, but that the time courses were different between the two structures.

Action Potentials↗
PubMed2004

Alteration of neuronal firing properties after in vivo experience in a FosGFP transgenic mouse.

Identifying the cells and circuits that underlie perception, behavior, and learning is a central goal of contemporary neuroscience. Although techniques such as lesion analysis, functional magnetic resonance imaging, 2-deoxyglucose studies, and induction of gene expression have been helpful in determining the brain areas responsible for particular functions, these methods are technically limited. Currently, there is no method that allows for the identification and electrophysiological characterization of individual neurons that are associated with a particular function in living tissue. We developed a strain of transgenic mice in which the expression of the green fluorescent protein (GFP) is controlled by the promoter of the activity-dependent gene c-fos. These mice enable an in vivo or ex vivo characterization of the cells and synapses that are activated by particular pharmacological and behavioral manipulations. Cortical and subcortical fosGFP expression could be induced in a regionally restricted manner after specific activation of neuronal ensembles. Using the fosGFP mice to identify discrete cortical areas, we found that neurons in sensory-spared areas rapidly regulate action potential threshold and spike frequency to decrease excitability. This method will enhance our ability to study the way neuronal networks are activated and changed by both experience and pharmacological manipulations. In addition, because activated neurons can be functionally characterized, this tool may enable the development of better pharmaceuticals that directly affect the neurons involved in disease states.

Action Potentials↗
PubMed1985

Studying speech perception in adolescent school-age children by utilizing primary color perception.

Since both speech and color have perceptual ramifications in language, the present study was developed to study speech perception through color perception. With the recent advances in perception generally and color perception specifically, this nontraditional approach to studying speech perception appeared reasonable. The 12 consonants utilized in this study generated 144 pairs of nonsense CV-syllables. The consonant ensemble was selected because it accommodated a maximum number of phonological features with a minimum number of phonemes. The 46 subjects, who were in junior high school with an age range of 11 through 14 years, responded to each of the stimulus pairs by assigning (associating) to it one of the six primary colors. Because of the perceptual orderliness associated with subjects' judgments, the results indicated that color can be used to study speech perception. Specific findings included the retrieval of sameness, fronting (or place), and voicing.

Adolescent↗
PubMed1993

Object and head orientation effects on symmetry perception defined by shape from shading.

A grey disc which exhibits a vertical gradient of shading usually appears convex if lighter above and concave if lighter below. This phenomenon was investigated by Howard et al (1990) who varied both the shading axis relative to gravity and the orientation of the head. Their results indicated that head-centric or retinocentric coordinates determined the depth effect rather than gravitational axes. However, several possible problems with their study were noted, not the least of which was the possible intrusion of response rather than perceptual factors in the task they used. Here, we attempted to use an indirect measure of the perception of depth from shading; rather than asking subjects whether discs looked convex or concave, we constructed ensembles of shaded discs which, in terms of depth from shading, were or were not bilaterally symmetrical about an horizontal axis. These stimulus displays were briefly flashed to prohibit the intrusion of conscious assumptions about direction of light sources. Subjects were never asked whether any discs looked concave or convex, merely whether the set of discs was or was not depth symmetrical. Results were generally consistent with those of Howard et al and supported the conclusion that depth from shading is largely a low-level and automatic mechanism.

Adult↗
PubMed1997

From Levinthal to pathways to funnels.

While the classical view of protein folding kinetics relies on phenomenological models, and regards folding intermediates in a structural way, the new view emphasizes the ensemble nature of protein conformations. Although folding has sometimes been regarded as a linear sequence of events, the new view sees folding as parallel microscopic multi-pathway diffusion-like processes. While the classical view invoked pathways to solve the problem of searching for the needle in the haystack, the pathway idea was then seen as conflicting with Anfinsen's experiments showing that folding is pathway-independent (Levinthal's paradox). In contrast, the new view sees no inherent paradox because it eliminates the pathway idea: folding can funnel to a single stable state by multiple routes in conformational space. The general energy landscape picture provides a conceptual framework for understanding both two-state and multi-state folding kinetics. Better tests of these ideas will come when new experiments become available for measuring not just averages of structural observables, but also correlations among their fluctuations. At that point we hope to learn much more about the real shapes of protein folding landscapes.

Kinetics↗
PubMed1990

Dimensional overlap: cognitive basis for stimulus-response compatibility--a model and taxonomy.

The classic problem of stimulus-response (S-R) compatibility (SRC) is addressed. A cognitive model is proposed that views the stimulus and response sets in S-R ensembles as categories with dimensions that may or may not overlap. If they do overlap, the task may be compatible or incompatible, depending on the assigned S-R mapping. If they do not overlap, the task is noncompatible regardless of the assigned mapping. The overlapping dimensions may be relevant or not. The model provides a systematic account of SRC effects, a taxonomy of simple performance tasks that were hitherto thought to be unrelated, and suggestive parallels between these tasks and the experimental paradigms that have traditionally been used to study attentional, controlled, and automatic processes.

Arousal↗
PubMed2006

Comparison of two Simon tasks: neuronal correlates of conflict resolution based on coherent motion perception.

The present study aimed at characterizing the neural correlates of conflict resolution in two variations of the Simon effect. We introduced two different Simon tasks where subjects had to identify shapes on the basis of form-from-motion perception (FFMo) within a randomly moving dot field, while (1) motion direction (motion-based Simon task) or (2) stimulus location (location-based Simon task) had to be ignored. Behavioral data revealed that both types of Simon tasks induced highly significant interference effects. Using event-related fMRI, we could demonstrate that both tasks share a common cluster of activated brain regions during conflict resolution (pre-supplementary motor area (pre-SMA), superior parietal lobule (SPL), and cuneus) but also show task-specific activation patterns (left superior temporal cortex in the motion-based, and the left fusiform gyrus in the location-based Simon task). Although motion-based and location-based Simon tasks are conceptually very similar (Type 3 stimulus-response ensembles according to the taxonomy of [Kornblum, S., Stevens, G. (2002). Sequential effects of dimensional overlap: findings and issues. In: Prinz, W., Hommel., B. (Eds.), Common mechanism in perception and action. Oxford University Press, Oxford, pp. 9-54]) conflict resolution in both tasks results in the activation of different task-specific regions probably related to the different sources of task-irrelevant information. Furthermore, the present data give evidence those task-specific regions are most likely to detect the relationship between task-relevant and task-irrelevant information.

Adult↗
PubMed2003

Genomics and variation of ionotropic glutamate receptors.

Sequencing of the human, mouse, and rat genomes has enabled a comprehensive informatics approach to gene families. This approach is informative for identification of new members of gene families, for cross-species sequence conservation related to functional conservation, for within-species diversity related to functional variation, and for historical effects of selection. This genome informatics approach also focuses our attention on genes whose genomic locations coincide with linkages to phenotypes. We are identifying ionotropic glutamate receptor (IGR) sequence variation by resequencing technologies, including denaturing high-performance liquid chromoatography (dHPLC), for screening and direct sequencing, and by information mining of public (e.g., dbSNP and ENSEMBL) and private (i.e., Celera Discovery System) sequence databases. Each of the 16 known IGRs is represented in these databases, their positions on a canonical physical map (for example, the Celera map) are established, and comparison to mouse and rat sequences has been performed, revealing substantial conservation of these genes, which are located on different chromosomes but found within syntenic groups of genes. A collection of 38 missense variants were identified by the informatics and resequencing approaches in several of these receptor genes, including GRIN2B, GRIN3B, GRIA2, GRIA3, and GRIK1. This represents only a fraction of the sequence variation across these genes, but, in fact, these may constitute a large fraction of the common polymorphisms at these genes, and these polymorphisms are a starting point for understanding the role of these receptors in neurogenetic variation. Genetically influenced human neurobehavioral phenotypes that are likely to be linked to IGR genetic variants include addictions, anxiety/dysphoria disorders, post-brain injury behavioral disorders, schizophrenia, epilepsy, pain perception, learning, and cognition. Thus, the effects of glutamate receptor variation may be protean, and the task of relating variation to behavior difficult. However, functional variants of (1) catechol-O-methyltransferase, (2) serotonin transporter, and (3) brain-derived neurotrophic factor have recently been linked both to behavioral differences and to intermediate phenotypes, suggesting a pathway by which functional variation at IGRs can be tied to an etiologically complex phenotype.

Animals↗
PubMed2002

Parallel processing of serial movements in prefrontal cortex.

A key idea in Lashley's formulation of the problem of serial order in behavior is the postulated neural representation of all serial elements before the action begins. We studied this question by recording the activity of individual neurons simultaneously in small ensembles in prefrontal cortex while monkeys copied geometrical shapes shown on a screen. Monkeys drew the shapes as sequences of movement segments, and these segments were associated with distinct patterns of neuronal ensemble activity. Here we show that these patterns were present during the time preceding the actual drawing. The rank of the strength of representation of a segment in the neuronal population during this time, as assessed by discriminant analysis, predicted the serial position of the segment in the motor sequence. An analysis of errors in copying and their neural correlates supplied additional evidence for this code and provided a neural basis for Lashley's hypothesis that errors in motor sequences would be most likely to occur when executing elements that had prior representations of nearly equal strength.

Animals↗