Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Probability 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 775 records · Page 43Linked to original sources
PubMed1995

Human chromosome classification using multilayer perceptron neural network.

A multilayer perceptron (MLP) neural network (NN) has been studied for human chromosome classification. Only 10-20 examples were required for the MLP NN to reach its ultimate performance classifying chromosomes of 5 types. The empirical dependence of the entropic error on the number of examples was found to be highly comparable to the 1/t function. The principal component analysis (PCA) was used, both for network initialization and for feature reduction purposes. The PCA demonstrated the importance of retaining most of the image information whenever small training sets are used. The MLP NN classifier outperformed the Bayes piecewise classifier for all the cases tested. The MLP classifier was found to be almost unsusceptible to the ratio of the number of training vectors to the number of features, whereas the piecewise classifier was highly dependent on this ratio.

Algorithms↗
PubMed1996

Hardware prototypes of a Boolean neural network and the simulated annealing optimization method.

Boolean Neural Network is a neural network that operates with binary weight values of "1" and "0". Otherwise it is formally analogous to the Multilayer Perceptron (MLP). Simulated Annealing is a stochastic optimization methods that is suitable for performing nonlinear multivariable optimization tasks. Training a Boolean Neural Network is a well-suited problem to this algorithm. However, the Simulated Annealing method is computationally heavy, which makes the training procedure slow. The training speed can be improved by using custom designed hardware for the whole system including the optimization method and the neural network. Hardware prototypes of a Boolean Neural Network and the Simulated Annealing optimization method have been designed using discrete components. The Boolean Neural Network implementation is basically a dynamically configurable feedforward network of Boolean logic gates of two inputs. The Simulated Annealing implementation is a general purpose hardware tool for multivariable optimization tasks. Here it is applied to do supervised training of the Boolean Neural Network hardware.

Algorithms↗
PubMed1996

An adaptive Boolean automation to model circadian cycles.

We propose a Boolean cellular automation to model an artificial adaptive living organism in order to investigate the development of cyclic vital functions during a simulated evolutionary process. The organism is endowed with a basic architecture consisting of several sensor (input), motor (output) and processing Boolean gates whose connectivity pattern is adapted with a genetic algorithm. Cyclic searching behaviors develop that are tuned to the spatial distribution of "food". Under additional assumptions we also find that internal pacemakers can develop to adapt plastically to the alternance of "light" an "darkness". These pacemakers coexist with a "free running" regime in which the circadian cycles persist and even in the absence of external periodic stimuli.

Adaptation, Physiological↗
PubMed1996

Constructive training methods for feedforward neural networks with binary weights.

Quantization of the parameters of a Perceptron is a central problem in hardware implementation of neural networks using a numerical technology. A neural model with each weight limited to a small integer range will require little surface of silicon. Moreover, according to Occam's razor principle, better generalization abilities can be expected from a simpler computational model. The price to pay for these benefits lies in the difficulty to train these kind of networks. This paper proposes essentially two new ideas for constructive training algorithms, and demonstrates their efficiency for the generation of feedforward networks composed of Boolean threshold gates with discrete weights. A proof of the convergence of these algorithms is given. Some numerical experiments have been carried out and the results are presented in terms of the size of the generated networks and of their generalization abilities.

Algorithms↗
PubMed1996

Nonlinear time series analysis by neural networks: a case study.

This paper presents a neural network approach to time-series analysis of a univariate nonlinear system. Feedforward networks are studied, and an appropriate network size is determined by different criteria computed on the basis of the performance of the models on the training and test sets. The analysis and conclusions drawn are supported by studies of the phase portraits of the models. By a proper choice of network size, the problems of over-parameterization are demonstrated to be avoided. The overfitting observed for larger networks is analyzed and the underlying reasons for their worse generalization capabilities are explained. Finally, some observations are made on the approximation provided by an oversized network with weights determined by an incomplete (interrupted) training and that of the optimal-sized network.

Cybernetics↗
PubMed2005

Spike timing in CA3 pyramidal cells during behavior: implications for synaptic transmission.

Spike timing is thought to be an important mechanism for transmitting information in the CNS. Recent studies have emphasized millisecond precision in spike timing to allow temporal summation of rapid synaptic signals. However, spike timing over slower time scales could also be important, through mechanisms including activity-dependent synaptic plasticity or temporal summation of slow postsynaptic potentials (PSPs) such as those mediated by kainate receptors. To determine the extent to which these slower mechanisms contribute to information processing, it is first necessary to understand the properties of behaviorally relevant spike timing over this slow time scale. In this study, we examine the activity of CA3 pyramidal cells during the performance of a complex behavioral task in rats. Sustained firing rates vary over a wide range, and the firing rate of a cell is poorly correlated with the behavioral cues to which the cell responds. Nonrandom interactions between successive spikes can last for several seconds, but the nonrandom distribution of interspike intervals (ISIs) can account for the majority of nonrandom multi-spike patterns. During a stimulus, cellular responses are temporally complex, causing a shift in spike timing that favors intermediate ISIs over short and long ISIs. Response discrimination between related stimuli occurs through changes in both response time-course and response intensity. Precise synchrony between cells is limited, but loosely correlated firing between cells is common. This study indicates that spike timing is regulated over long time scales and suggests that slow synaptic mechanisms could play a substantial role in information processing in the CNS.

Action Potentials↗
PubMed1994

Differences in P3 amplitudes between schizophrenics and healthy controls vary between the different events presented in a guessing task.

P3 amplitudes were often found to be reduced in schizophrenics, but in varying degree. We studied in a guessing task whether variations of group differences could even be obtained within the same task, by measuring P3 in the potentials evoked by the three qualitatively different events that occurred in each trial. We hypothesized that such variations were due to variations of subjective task relevance associated with each event. In each trial, a light or a sound was presented. Subjects had to confirm this 'primary event' by a delayed response, and then the current amount of money earned by correct responses was displayed. In the certain condition, the primary event was preceded by the appropriate cue on the screen ('light' or 'sound') whereas in the uncertain condition, the word 'guess' appeared. The task-relevance hypothesis predicted that schizophrenics would have smaller P3s than the control group with the primary event in the uncertain condition, but that the groups would not differ for the P3s evoked by the other events (certain primary event, cue and earnings display in uncertain and certain conditions). Essentially, this predicted pattern of results was obtained, but additional assumptions are needed for the task-relevance hypothesis to account for the precise pattern of results. Analysis of subjects' guessing behavior showed that control subjects' guesses were affected by the outcome of their preceding guesses while schizophrenics' guesses were not. This result provides some additional support for the hypothesis that the guessing task is less relevant to schizophrenics than to control subjects.

Adult↗
PubMed1978

The elderly and the control of simple behaviour by probabilistic information.

The probabilities of events in the environment are important as controllers of how we cope with that environment. Attention is deployed to parts of the visual field where important events are most likely to happen; in predicting what will happen next in a situation, probable things tend to be predicted. This paper assess adult age differences in the use of probabilistic information, using the laboratory task-settings of choice reaction time, simple prediction, and selective attending. In almost all situations the elderly are less influenced by event probabilities than their juniors, although they are just as capable of registering the probabilities involved. This consistent age effect contradicts the predictions of Griew's theory which assumes that behavioural experience operates by increasing the use of probabilistic information. A second finding was that value and reward manipulations which alter the response patterns of the young do not influence the elderly to the same extent. It is tentatively concluded that attenuated use of probability information characterises the elderly, as does a reduced response to simple payoffs. Some avenues of theoretical exploration are discussed, together with some caveats concerning the applicability of the findings.

Adult↗
PubMed2000

Sequential monte carlo methods To train neural network models

We discuss a novel strategy for training neural networks using sequential Monte Carlo algorithms and propose a new hybrid gradient descent sampling importance resampling algorithm (HySIR). In terms of computational time and accuracy, the hybrid SIR is a clear improvement over conventional sequential Monte Carlo techniques. The new algorithm may be viewed as a global optimization strategy that allows us to learn the probability distributions of the network weights and outputs in a sequential framework. It is well suited to applications involving on-line, nonlinear, and nongaussian signal processing. We show how the new algorithm outperforms extended Kalman filter training on several problems. In particular, we address the problem of pricing option contracts, traded in financial markets. In this context, we are able to estimate the one-step-ahead probability density functions of the options prices.

Journal Article↗
PubMed1997

A penalty-function approach for pruning feedforward neural networks.

This article proposes the use of a penalty function for pruning feedforward neural network by weight elimination. The penalty function proposed consists of two terms. The first term is to discourage the use of unnecessary connections, and the second term is to prevent the weights of the connections from taking excessively large values. Simple criteria for eliminating weights from the network are also given. The effectiveness of this penalty function is tested on three well-known problems: the contiguity problem, the parity problems, and the monks problems. The resulting pruned networks obtained for many of these problems have fewer connections than previously reported in the literature.

Algorithms↗
PubMed2005

Lower prepulse inhibition in children with the 22q11 deletion syndrome.

OBJECTIVE: The 22q11 deletion syndrome is associated with a range of possible physical anomalies, probable ongoing learning disabilities, and a specific constellation of neuropsychological deficits, including impairments in selective and executive visual attention, working memory, and sensorimotor functioning. It has been estimated that 25% of the children with 22q11 deletion syndrome go on to develop schizophrenia in late adolescence or adulthood. This is of urgent concern. Specification of early brain network vulnerabilities may provide a basis for early intervention while indicating critical links between genes and severe psychiatric illness. Neuropsychological studies of children with 22q11 deletion syndrome have implicated an array of potentially aberrant brain pathways. This study was conducted to determine whether preattentive processing ("sensorimotor gating") deficits are present in this population. METHOD: The authors administered a test of prepulse inhibition to 25 children with 22q11 deletion syndrome and their 23 sibling comparison subjects, ages 6-13. It was predicted that the children with 22q11 deletion syndrome would have lower prepulse inhibition than the comparison subjects. RESULTS: Prepulse inhibition in the children with 22q11 deletion syndrome (26.06%) was significantly less than that of the sibling comparison subjects (46.41%). Secondary analyses suggested that this decrement did not reflect developmental delay, and lower prepulse inhibition was associated with particular subsyndromal symptoms in some children. CONCLUSIONS: Sensorimotor gating is lower in children with 22q11 deletion syndrome. These findings may indicate specific brain circuits that are anomalous in 22q11 deletion syndrome.

Abnormalities, Multiple↗
PubMed1989

Fuzzy logical approach to perception of dot numerosity.

The present study utilized a fuzzy logical approach for understanding human perception or judgments of dot numerosity. In Exp. 1 subjects were required to view dot patterns and to judge the truthfulness of the single and combined statements which asserted that the number of dots was large. The results indicated that (a) the rules based on the minimum and maximum truthfulness of the component statements best approximate subjective conjunction and disjunction about dot numerosity, when subjects kept the operations of the standard logic system in mind. (b) When the subjects based their judgments on perceptive impression, their judgments were best fitted by the multiplicative form.

Adult↗
PubMed2006

The "ripple effect": cultural differences in perceptions of the consequences of events.

Previous research has demonstrated that people from East Asian cultural backgrounds make broader, more complex causal attributions than do people from Western cultural backgrounds. In the current research, the authors hypothesized that East Asians also would be aware of a broader, more complex distribution of consequences of events. Four studies assessed cultural differences in perceptions of the consequences of (a) a shot in a game of pool, (b) an area being converted into a national park, (c) a chief executive officer firing employees, and (d) a car accident. Across all four studies, compared to participants from Western cultural backgrounds, participants from East Asian cultural backgrounds were more aware of the indirect, distal consequences of events. This pattern occurred on a variety of measures, including spontaneously generated consequences, estimations of an event's impact on subsequent events, perceived responsibility, and predicted affective reactions. Implications for our understanding of cross-cultural psychology and social perception are discussed.

Adolescent↗