PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “Neural Networks, Computer”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,135 records · Page 63Linked to original sources
PubMed1996

First use of cognitive algorithms in investigations under compensated gravity.

In the present paper the use of cognitive algorithms for solving a wide spectrum of problems which often arise in investigations under compensated gravity is suggested. Applying such algorithms in the preparation and performance of experiments provides a substantial assistance to the experimentator as the behaviour of complex processes can be described and predicted correctly even when unexpected perturbations occur. Furthermore, an essential advantage of cognitive computing consists in the fact that the description and optimisation of the processes considered are possible also in such cases in which the corresponding basic equations are not known or not treatable practically. For convenience, the basic ideas of cognitive algorithms are discussed here. Due to their special relevance for investigations under compensated gravity algorithms based on fuzzy logic (FL) and artificial neuronal networks (ANN) are elucidated more in detail. In order to illustrate some advantages of cognitive computing exemplary results for the flow field induced by coaxial rotating disks are given. This represents the first attempt to use the benefits provided by cognitive algorithms in investigations under compensated gravity. The flow field between rotating disks plays an important role not only in experiments under compensated gravity but also in a wide range of terrestrial applications. A comparison of the results found by solving the Navier-Stokes equations and those from the prediction performed by ANN adequately trained shows an excellent agreement. However, the calculation times needed by the ANN are significantly smaller than that of the direct numerical simulation. Therefore, the real time prediction of the results from a running experiment seems to be possible.

Algorithms↗
PubMed2006

Application of generalized dynamic neural networks to biomedical data.

This paper reviews the application of continuous recurrent neural networks with time-varying weights to pattern recognition tasks in medicine. A general learning algorithm based on Pontryagin's maximum principle is recapitulated, and possibilities of improving the generalization capabilities of these networks are given. The effectiveness of the methods is demonstrated by three different real-world examples taken from the fields of anesthesiology, orthopedics, and radiology.

Algorithms↗
PubMed2002

Real-time computing without stable states: a new framework for neural computation based on perturbations.

A key challenge for neural modeling is to explain how a continuous stream of multimodal input from a rapidly changing environment can be processed by stereotypical recurrent circuits of integrate-and-fire neurons in real time. We propose a new computational model for real-time computing on time-varying input that provides an alternative to paradigms based on Turing machines or attractor neural networks. It does not require a task-dependent construction of neural circuits. Instead, it is based on principles of high-dimensional dynamical systems in combination with statistical learning theory and can be implemented on generic evolved or found recurrent circuitry. It is shown that the inherent transient dynamics of the high-dimensional dynamical system formed by a sufficiently large and heterogeneous neural circuit may serve as universal analog fading memory. Readout neurons can learn to extract in real time from the current state of such recurrent neural circuit information about current and past inputs that may be needed for diverse tasks. Stable internal states are not required for giving a stable output, since transient internal states can be transformed by readout neurons into stable target outputs due to the high dimensionality of the dynamical system. Our approach is based on a rigorous computational model, the liquid state machine, that, unlike Turing machines, does not require sequential transitions between well-defined discrete internal states. It is supported, as the Turing machine is, by rigorous mathematical results that predict universal computational power under idealized conditions, but for the biologically more realistic scenario of real-time processing of time-varying inputs. Our approach provides new perspectives for the interpretation of neural coding, the design of experiments and data analysis in neurophysiology, and the solution of problems in robotics and neurotechnology.

Action Potentials↗
PubMed1992

Neural computing in cancer drug development: predicting mechanism of action.

Described here are neural networks capable of predicting a drug's mechanism of action from its pattern of activity against a panel of 60 malignant cell lines in the National Cancer Institute's drug screening program. Given six possible classes of mechanism, the network misses the correct category for only 12 out of 141 agents (8.5 percent), whereas linear discriminant analysis, a standard statistical technique, misses 20 out of 141 (14.2 percent). The success of the neural net indicates several things. (i) The cell line response patterns are rich in information about mechanism. (ii) Appropriately designed neural networks can make effective use of that information. (iii) Trained networks can be used to classify prospectively the more than 10,000 agents per year tested by the screening program. Related networks, in combination with classical statistical tools, will help in a variety of ways to move new anticancer agents through the pipeline from in vitro studies to clinical application.

Alkylating Agents↗
PubMed1999

Sequential RAM-based neural networks: learnability, generalisation, knowledge extraction, and grammatical inference.

A fundamental question in the field of artificial neural networks is what set of problems a given class of networks can perform (computability). Such a problem can be made less general, but no less important, by asking what these networks could learn by using a given training procedure (learnability). The basic purpose of this paper is to address the learnability problem. Specifically, it analyses the learnability of sequential RAM-based neural networks. The analytical tools used are those of Automata Theory. In this context, this paper establishes which class of problems and under what conditions such networks, together with their existing learning rules, can learn and generalize. This analysis also yields techniques for both extracting knowledge from and inserting knowledge into the networks. The results presented here, besides helping in a better understanding of the temporal behaviour of sequential RAM-based networks, could also provide useful insights for the integration of the symbolic/connectionist paradigms.

Algorithms↗
PubMed1997

An analog memory circuit for spiking silicon neurons.

A simple circuit is described that functions as an analog memory whose state and dynamics are directly controlled by pulsatile inputs. The circuit has been incorporated into a silicon neuron with a spatially extensive dendritic tree as a means of controlling the spike firing threshold of an integrate-and-fire soma. Spiking activity generated by the neuron itself and by other units in a network can thereby regulate the neuron's excitability over time periods ranging from milliseconds to many minutes. Experimental results are presented showing applications to temporal edge sharpening, bistable behavior, and a network that learns in the manner of classical conditioning.

Action Potentials↗
PubMed1997

Symmetry breaking and training from incomplete data with Radial Basis Boltzmann Machines.

A Radial Basis Boltzmann Machine (RBBM) is a specialized Boltzmann Machine architecture that combines feed-forward mapping with probability estimation in the input space, and for which very efficient learning rules exist. The hidden representation of the network displays symmetry breaking as a function of the noise in the dynamics. Thus, generalization can be studied as a function of the noise in the neuron dynamics instead of as a function of the number of hidden units. We show that the RBBM can be seen as an elegant alternative of k-nearest neighbor, leading to comparable performance without the need to store all data. We show that the RBBM has good classification performance compared to the MLP. The main advantage of the RBBM is that simultaneously with the input-output mapping, a model of the input space is obtained which can be used for learning with missing values. We derive learning rules for the case of incomplete data, and show that they perform better on incomplete data than the traditional learning rules on a 'repaired' data set.

Computer Simulation↗
PubMed1997

Comparative study of artificial neural networks in the discrimination between benign from malignant gastric cells.

OBJECTIVE: To compare the accuracy of two different artificial neural networks (ANNs) for the discrimination of benign and malignant gastric lesions using morphometric and textural data on the nucleus. STUDY DESIGN: Three thousand cells from 30 cancer cases, 26 cases of gastritis and 64 cases of ulcer were selected as a training set, and an additional 10,300 cells from equal cases of cancer, gastritis and ulcer were used as a test set using two different neural net architectures: back propagation (BP) and learning vector quantizer (LVQ). Images of routinely processed gastric smears stained by the Papanicolaou technique were processed by a custom image analysis system. RESULTS: Application of the BP and three variations of the LVQ established correct classification of more than 97% of the benign cells and more than 95% of the malignant cells, obtaining an overall accuracy of more than 97%. CONCLUSION: This study not only presents a comparative study of the abilities of ANNs but also indicates that the use of ANNs and image morphometry may offer useful information on the potential of malignancy of gastric cells.

Adult↗
PubMed2001

Multialternative decision field theory: a dynamic connectionist model of decision making.

The authors interpret decision field theory (J. R. Busemeyer & J. T. Townsend, 1993) as a connectionist network and extend it to accommodate multialternative preferential choice situations. This article shows that the classic weighted additive utility model (see R. L. Keeney & H. Raiffa, 1976) and the classic Thurstone preferential choice model (see L. L. Thurstone, 1959) are special cases of this new multialternative decision field theory (MDFT), which also can emulate the search process of the popular elimination by aspects (EBA) model (see A. Tversky, 1969). The new theory is unique in its ability to explain several central empirical results found in the multialternative preference literature with a common set of principles. These empirical results include the similarity effect, the attraction effect, and the compromise effect, and the complex interactions among these three effects. The dynamic nature of the model also implies strong testable predictions concerning the moderating effect of time pressure on these three effects.

Artificial Intelligence↗
PubMed2002

Prediction of the progression of thyroid-associated ophthalmopathy at first ophthalmologic examination: use of a neural network.

In the present work we analyzed patients with thyroid-associated ophthalmopathy (TAO) at various clinical stages of disease progression and implemented a model of neural analysis for disease classification and prediction of progression. We studied 246 patients (group 1), seen only once because they had absent, minimal, or inactive TAO and 152 patients (group 2), seen two or more times because of active and/or progressive TAO. The ophthalmologic assessment included: (1) lid fissure measurement; (2) Hertel; (3) color vision; (4) cover test and Hess screen; (5) visual acuity; (6) tonometry; (7) fundus examination; (8) visual field; (9) orbital computed tomography (CT) scan or ultrasound. A back propagation model of neural network was based on the relative variations of 13 clinical eye signs (input variables) for classification and prediction of disease progression (output variable). Approximately 300 eyes (20%) were randomly selected as a test group. Correlation between expected and calculated patients' classification was highly significant (p < 0.00001). Concordance between clinical assessment and the neural network prediction was obtained in 78 of 117 eyes (67%). We have developed a neural model that allows classification of TAO and preliminary prediction of disease progression at the first clinical examination. The results are validating the classification into the two groups on which our initial assumption was based.

Disease Progression↗
PubMed2006

Dimensional reduction for reward-based learning.

Reward-based learning in neural systems is challenging because a large number of parameters that affect network function must be optimized solely on the basis of a reward signal that indicates improved performance. Searching the parameter space for an optimal solution is particularly difficult if the network is large. We show that Hebbian forms of synaptic plasticity applied to synapses between a supervisor circuit and the network it is controlling can effectively reduce the dimension of the space of parameters being searched to support efficient reinforcement-based learning in large networks. The critical element is that the connections between the supervisor units and the network must be reciprocal. Once the appropriate connections have been set up by Hebbian plasticity, a reinforcement-based learning procedure leads to rapid learning in a function approximation task. Hebbian plasticity within the network being supervised ultimately allows the network to perform the task without input from the supervisor.

Animals↗
PubMed2002

Long-range synchronization of gamma and beta oscillations and the plasticity of excitatory and inhibitory synapses: a network model.

The ability of oscillating networks to synchronize despite significant separation in space, and thus time, is of biological significance, given that human gamma activity can synchronize over distances of several millimeters to centimeters during perceptual and learning tasks. We use computer simulations of networks consisting of excitatory pyramidal cells (e-cells) and inhibitory interneurons (i-cells), modeling two tonically driven assemblies separated by large (>or=8 ms) conduction delays. The results are as follows. 1) Two assemblies separated by large conduction delays can fire synchronously at beta frequency (with i-cells firing at gamma frequency) under two timing conditions: e-cells of (say) assembly 2 are still inhibited "delay + spike generation milliseconds" after the e-cell beat of assembly 1; this means that the e-cell inhibitory postsynaptic potential (IPSP) cannot be significantly shorter than the delay (2-site effect). This implies for a given decay time constant that the interneuron --> pyramidal cell conductances must be large enough. The e-cell IPSP must last longer than the i-cell IPSP, i.e., the interneuron --> pyramidal cell conductance must be sufficiently large and the interneuron --> interneuron conductance sufficiently small (local effect). 2) We define a "long-interval doublet" as a pair of interneuron action potentials-separated by approximately "delay milliseconds"-in which a) the first spike is induced by tonic inputs and/or excitation from nearby e-cells, while b) the second spike is induced by (delayed) excitation from distant e-cells. "Long-interval population doublets" (long-interval doublets of the i-cell population) are necessary for synchronized firing in our networks. Failure to produce them leads to almost anti-phase activity at gamma frequency. 3) An (almost) anti-phase oscillation is the most stable oscillation pattern of two assemblies that are separated by axonal conduction delays of approximately one-half a gamma period (delays from 8 to 17 ms in our simulations) and that are firing at gamma frequency. 4) Two assemblies separated by large conduction delays can synchronize their activity with the help of interneuron plasticity. They can also synchronize without pyramidal cell --> pyramidal cell connections being present. The presence of pyramidal cell --> pyramidal cell connections allows, however, for synchronization if other parameters are at inappropriate values for synchronization to occur. 5) Synchronization of two assemblies separated by large conduction delays with the help of interneuron plasticity is not simply due to slowing down of the oscillation frequency. It is reached with the help of a "synchronizing-weak-beat," which induces sudden changes in the oscillation period length of the two assemblies.

Action Potentials↗
PubMed2006

An integrate-and-fire model for synchronized bursting in a network of cultured cortical neurons.

It has been suggested that spontaneous synchronous neuronal activity is an essential step in the formation of functional networks in the central nervous system. The key features of this type of activity consist of bursts of action potentials with associated spikes of elevated cytoplasmic calcium. These features are also observed in networks of rat cortical neurons that have been formed in culture. Experimental studies of these cultured networks have led to several hypotheses for the mechanisms underlying the observed synchronized oscillations. In this paper, bursting integrate-and-fire type mathematical models for regular spiking (RS) and intrinsic bursting (IB) neurons are introduced and incorporated through a small-world connection scheme into a two-dimensional excitatory network similar to those in the cultured network. This computer model exhibits spontaneous synchronous activity through mechanisms similar to those hypothesized for the cultured experimental networks. Traces of the membrane potential and cytoplasmic calcium from the model closely match those obtained from experiments. We also consider the impact on network behavior of the IB neurons, the geometry and the small world connection scheme.

Action Potentials↗
PubMed2004

Adaptive hybrid control for linear piezoelectric ceramic motor drive using diagonal recurrent CMAC network.

This paper presents an adaptive hybrid control system using a diagonal recurrent cerebellar-model-articulation-computer (DRCMAC) network to control a linear piezoelectric ceramic motor (LPCM) driven by a two-inductance two-capacitance (LLCC) resonant inverter. Since the dynamic characteristics and motor parameters of the LPCM are highly nonlinear and time varying, an adaptive hybrid control system is therefore designed based on a hypothetical dynamic model to achieve high-precision position control. The architecture of DRCMAC network is a modified model of a cerebellar-model-articulation-computer (CMAC) network to attain a small number of receptive-fields. The novel idea of this study is that it employs the concept of diagonal recurrent neural network (DRNN) in order to capture the system dynamics and convert the static CMAC into a dynamic one. This adaptive hybrid control system is composed of two parts. One is a DRCMAC network controller that is used to mimic a conventional computed torque control law due to unknown system dynamics, and the other is a compensated controller with bound estimation algorithm that is utilized to recover the residual approximation error for guaranteeing the stable characteristic. The effectiveness of the proposed driving circuit and control system is verified with hardware experiments under the occurrence of uncertainties. In addition, the advantages of the proposed control scheme are indicated in comparison with a traditional integral-proportional (IP) position control system.

Algorithms↗
PubMed2003

Speaker verification using committee neural networks.

Security is a major problem in web based access or remote access to data bases. In the present study, the technique of committee neural networks was developed for speech based speaker verification. Speech data from the designated speaker and several imposters were obtained. Several parameters were extracted in the time and frequency domains, and fed to neural networks. Several neural networks were trained and the five best performing networks were recruited into the committee. The committee decision was based on majority voting of the member networks. The committee opinion was evaluated with further testing data. The committee correctly identified the designated speaker in (50 out of 50) 100% of the cases and rejected imposters in (150 out of 150) 100% of the cases. The committee decision was not unanimous in majority of the cases tested.

Algorithms↗
PubMed2005

A brain-like neural network for periodicity analysis.

This paper introduces a brain-like neural model for sound processing. The periodicity analyzing network (PAN) is a bio-inspired neural network of spiking neurons. The PAN consists of complex models of neurons, which can be used for understanding the dynamics of individual neurons and neuronal networks. On a technical level, the PAN is able to compute the ratio of modulation and carrier frequency of harmonic sound signals. The PAN model may, therefore, be used in audio signal processing applications, such as sound source separation, periodicity analysis, and the cocktail party problem.

Algorithms↗
PubMed1993

A model for neural representation of binocular disparity in striate cortex: distributed representation and veto mechanism.

A model in striate cortex is proposed for a distributed neural representation of binocular disparity with a simple cell. In the model, disparity is represented by "far", "near" and "tuned inhibitory" simple cells. However, the representation will be vetoed by model cells where disparity is excessively large. The veto mechanism consists of a neural network of the model cell which received output from simple cells and which interacts with neighbors. The mechanism is necessary, the model cell responds like a simple cell, and the network is physiologically plausible in the brain. Computer simulation on the neural network model with random dot stereography indicates reasonable performance.

Computer Simulation↗