SEARCH · PubMed Health

Results for “Multilayer Perceptrons”

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 37 records · Page 2Linked to original sources

PubMed1998

The inverted multilayer perceptron as a consistency measuring device in neuroanatomical image analysis.

This protocol describes the use of optical imaging and artificial neural networks to analyze the complex patterns of transsynaptic transfer of WGA-HRP from olfactory bulbar axon terminals to olfactory tubercle projection neurons. This constitutes a unique application of neural networks to pattern recognition in a neuroanatomical tract-tracing study. Furthermore, the type of neural network used in this study, i.e., the inverted multilayer perceptron, possesses a novel construction that suits it for measuring consistency in data clusters. In the experiment described, the perceptron functioned as a consistency measuring device that revealed nonlinear relationships between the site of bulbar injection and the degree of regional labeling in the olfactory tubercle. Nonparametric statistical tests were used concurrently to validate the results of this previously untested form of the multilayer perceptron.

Animals↗

PubMed2000

Efficient block training of multilayer perceptrons.

The attractive possibility of applying layerwise block training algorithms to multilayer perceptrons MLP, which offers initial advantages in computational effort, is refined in this article by means of introducing a sensitivity correction factor in the formulation. This results in a clear performance advantage, which we verify in several applications. The reasons for this advantage are discussed and related to implicit relations with second-order techniques, natural gradient formulations through Fisher's information matrix, and sample selection. Extensions to recurrent networks and other research lines are suggested at the close of the article.

Algorithms↗

PubMed1996

Vowel-onset detection by vowel-strength measurement, cochlear-nucleus simulation, and multilayer perceptrons.

An algorithm for detection of vowel onsets in fluent speech was presented by Hermes [j. Acoust. Soc. Am. 87, 866-873 (1990)]. Performance tests showed that detection was good for fluent speech, although the parameter settings had to be modified for application to well-articulated speech. One of the purposes of the algorithm was application to speech by deaf persons, for which it failed completely. In order to improve the algorithm and to make it more generally applicable, two alternative detection strategies have been explored in the present study. These strategies were (a) simulation of transient-chopper responses in the cochlear nucleus and (B) training of multilayer perceptrons. Two large databases of read speech have been used for performance comparison of the original algorithm and the new strategies. The strategy based on simulating cochlear-nucleus responses is found both to result in a higher false-alarm rate than the original algorithm and to be rather level dependent. On the other hand, the performance of a multilayer-perceptron network, trained on mel-scaled spectra, is comparable to the performance of the Hermes algorithm. In more general terms, the results suggest that temporal information on intensity and (rough) spectral envelope are important for human vowel-onset detection behavior. In formation on harmonicity can be used as a secondary source of information to avoid detection of mainly unvoiced, nonvowel onsets.

Algorithms↗

PubMed1997

Estimates of Storage Capacity of Multilayer Perceptron with Threshold Logic Hidden Units.

We estimate the storage capacity of multilayer perceptron with n inputs, h(1) threshold logic units in the first hidden layer and 1 output. We show that if the network can memorize 50% of all dichotomies of a randomly selected N-tuple of points of R(n) with probability 1, then N</=2(nh(1)+1), while at 100% memorization N</=nh(1)+1. Furthermore, if the bounds are reached, then the first hidden layer must be fully connected to the input. It is shown that such a network has memory capacity (in the sense of Cover) between nh(1)+1 and 2(nh(1)+1) input patterns and for the most efficient networks in this class between 1 and 2 input patterns per connection. Comparing these results with the recent estimates of VC-dimension we find that in contrast to a single neuron case, the VC-dimension exceeds the capacity for a sufficiently large n and h(1). The results are based on the derivation of an explicit expression for the number of dichotomies which can be implemented by such a network for a special class of N-tuples of input patterns which has a positive probability of being randomly chosen.

Journal Article↗

PubMed2006

CATSMLP: toward a robust and interpretable multilayer perceptron with sigmoid activation functions.

Enhancing the robustness and interpretability of a multilayer perceptron (MLP) with a sigmoid activation function is a challenging topic. As a particular MLP, additive TS-type MLP (ATSMLP) can be interpreted based on single-stage fuzzy IF-THEN rules, but its robustness will be degraded with the increase in the number of intermediate layers. This paper presents a new MLP model called cascaded ATSMLP (CATSMLP), where the ATSMLPs are organized in a cascaded way. The proposed CATSMLP is a universal approximator and is also proven to be functionally equivalent to a fuzzy inference system based on syllogistic fuzzy reasoning. Therefore, the CATSMLP may be interpreted based on syllogistic fuzzy reasoning in a theoretical sense. Meanwhile, due to the fact that syllogistic fuzzy reasoning has distinctive advantage over single-stage IF-THEN fuzzy reasoning in robustness, this paper proves in an indirect way that the CATSMLP is more robust than the ATSMLP in an upper-bound sense. Several experiments were conducted to confirm such a claim.

Journal Article↗

PubMed1995

Methods of training and constructing multilayer perceptrons with arbitrary pattern sets.

This paper presents two compensation methods for multilayer perceptrons (MLPs) which are very difficult to train by traditional Back Propagation (BP) methods. For MLPs trapped in local minima, compensating methods can correct the wrong outputs one by one using constructing techniques until all outputs are right, so that the MLPs can skip from the local minima to the global minima. A hidden neuron is added as compensation for a binary input three-layer perceptron trapped in a local minimum; and one or two hidden neurons are added as compensation for a real input three-layer perception. For a perceptron of more than three layers, the second hidden layer from behind will be temporarily treated as the input layer during compensation, hence the above methods can also be used. Examples are given.

Algorithms↗

PubMed2004

Avoiding overfitting in multilayer perceptrons with feeling-of-knowing using self-organizing maps.

Overfitting in multilayer perceptron (MLP) training is a serious problem. The purpose of this study is to avoid overfitting in on-line learning. To overcome the overfitting problem, we have investigated feeling-of-knowing (FOK) using self-organizing maps (SOMs). We propose MLPs with FOK using the SOMs method to overcome the overfitting problem. In this method, the learning process advances according to the degree of FOK calculated using SOMs. The mean square error obtained for the test set using the proposed method is significantly less than that in a conventional MLP method. Consequently, the proposed method avoids overfitting.

Algorithms↗

PubMed1999

Derivation of the multilayer perceptron weight constraints for direct network interpretation and knowledge discovery.

This paper examines the multilayer perceptron (MLP) network from a hidden layer decision region perspective and derives the output layer and hidden layer weight constraints that the network must satisfy in performing a general classification task. This provides a foundation for direct knowledge discovery from the MLP, using a new method published by the author, which finds the key inputs that the MLP uses to classify an input case. The knowledge that the MLP network learns from the training examples is represented as ranked data relationships and induced rules, which can be used to validate the MLP network. The bounds of the network knowledge are established in the n-dimensional input space and a measure of the limit of the MLP network knowledge is proposed. An algorithm is presented for the calculation of the maximum number of hidden layer decision regions in the MLP input space.

Journal Article↗

PubMed2001

Specification of training sets and the number of hidden neurons for multilayer perceptrons.

This work concerns the selection of input-output pairs for improved training of multilayer perceptrons, in the context of approximation of univariate real functions. A criterion for the choice of the number of neurons in the hidden layer is also provided. The main idea is based on the fact that Chebyshev polynomials can provide approximations to bounded functions up to a prescribed tolerance, and, in turn, a polynomial of a certain order can be fitted with a three-layer perceptron with a prescribed number of hidden neurons. The results are applied to a sensor identification example.

Algorithms↗

PubMed2002

Automatic discrimination between supraventricular and ventricular tachycardia using a multilayer perceptron in implantable cardioverter defibrillators.

The morphological analysis of implantable cardioverter defibrillator (ICD) stored electrograms (EGM) using a multilayer perceptron (MLP) has been proposed for discrimination between supraventricular and ventricular arrhythmias. However, a reliable estimation of the accuracy of MLP methods is lacking. The aim of the study was to compare the morphology and spectrum-based MLP with more conventional morphology-based algorithms in a large series of ICD-stored episodes of arrhythmia. One set of ICD-stored electrograms was used for control and training purposes and a second one, consisting of spontaneous episodes in patients with dual chamber ICDs, for validation of the MLP performance. The correlation waveform analysis (CWA) and the EGM width criterion were compared with MLP methods. Bootstrap resampling techniques were used to extract the relevant information in the MLP training. The morphology-based MLP achieved better discrimination than any other method, with areas under the receiver operating characteristic (ROC) curve (tolerance intervals): 0.96 (0.81, 0.96) for MLP, 0.91 (0.77, 0.94) for CWA, and 0.68 (0.49, 0.78) for EGM width in the validation set. A specificity of 73.0% was obtained at 95% sensitivity, compared with 38.1% and 55.1% using CWA and EGM width criteria, respectively. In contrast, the generalization capabilities of spectral-based MLP methods are poor, showing a lower area under the ROC curve in the validation set. Time-domain MLP techniques may be useful for the morphological analysis of the intracardiac EGM signal stored by ICD devices. When properly trained and validated, these methods perform better than other commonly used morphological criteria for discrimination between supraventricular and ventricular arrhythmias.

Algorithms↗

PubMed2000

Adaptive method of realizing natural gradient learning for multilayer perceptrons.

The natural gradient learning method is known to have ideal performances for on-line training of multilayer perceptrons. It avoids plateaus, which give rise to slow convergence of the backpropagation method. It is Fisher efficient, whereas the conventional method is not. However, for implementing the method, it is necessary to calculate the Fisher information matrix and its inverse, which is practically very difficult. This article proposes an adaptive method of directly obtaining the inverse of the Fisher information matrix. It generalizes the adaptive Gauss-Newton algorithms and provides a solid theoretical justification of them. Simulations show that the proposed adaptive method works very well for realizing natural gradient learning.

Algorithms↗

PubMed2002

Representation and extrapolation in multilayer perceptrons.

To give an adequate explanation of cognition and perform certain practical tasks, connectionist systems must be able to extrapolate. This work explores the relationship between input representation and extrapolation, using simulations of multilayer perceptrons trained to model the identity function. It has been discovered that representation has a marked effect on extrapolation.

Cognition↗

PubMed2003

A quantified sensitivity measure for multilayer perceptron to input perturbation.

The sensitivity of a neural network's output to its input perturbation is an important issue with both theoretical and practical values. In this article, we propose an approach to quantify the sensitivity of the most popular and general feedforward network: multilayer perceptron (MLP). The sensitivity measure is defined as the mathematical expectation of output deviation due to expected input deviation with respect to overall input patterns in a continuous interval. Based on the structural characteristics of the MLP, a bottom-up approach is adopted. A single neuron is considered first, and algorithms with approximately derived analytical expressions that are functions of expected input deviation are given for the computation of its sensitivity. Then another algorithm is given to compute the sensitivity of the entire MLP network. Computer simulations are used to verify the derived theoretical formulas. The agreement between theoretical and experimental results is quite good. The sensitivity measure can be used to evaluate the MLP's performance.

Evaluation Studies as Topic↗

PubMed2000

Models to predict cardiovascular risk: comparison of CART, multilayer perceptron and logistic regression.

The estimate of a multivariate risk is now required in guidelines for cardiovascular prevention. Limitations of existing statistical risk models lead to explore machine-learning methods. This study evaluates the implementation and performance of a decision tree (CART) and a multilayer perceptron (MLP) to predict cardiovascular risk from real data. The study population was randomly splitted in a learning set (n = 10,296) and a test set (n = 5,148). CART and the MLP were implemented at their best performance on the learning set and applied on the test set and compared to a logistic model. Implementation, explicative and discriminative performance criteria are considered, based on ROC analysis. Areas under ROC curves and their 95% confidence interval are 0.78 (0.75-0.81), 0.78 (0.75-0.80) and 0.76 (0.73-0.79) respectively for logistic regression, MLP and CART. Given their implementation and explicative characteristics, these methods can complement existing statistical models and contribute to the interpretation of risk.

Artificial Intelligence↗

PubMed2005

Detection of ophthalmic arterial doppler signals with Behcet disease using multilayer perceptron neural network.

Doppler ultrasound is known as a reliable technique, which demonstrates the flow characteristics and resistance of ophthalmic arteries. In this study, ophthalmic arterial Doppler signals were obtained from 106 subjects, 54 of whom suffered from ocular Behcet disease while the rest were healthy subjects. Multilayer perceptron neural network (MLPNN) employing delta-bar-delta training algorithm was used to detect the presence of ocular Behcet disease. Spectral analysis of the ophthalmic arterial Doppler signals was performed by least squares (LS) autoregressive (AR) method for determining the MLPNN inputs. The MLPNN was trained with training set, cross validated with cross validation set and tested with testing set. All these data sets were obtained from ophthalmic arteries of healthy subjects and subjects suffering from ocular Behcet disease. Performance indicators and statistical measures were used for evaluating the MLPNN. The correct classification rate was 96.43% for healthy subjects and 93.75% for unhealthy subjects suffering from ocular Behcet disease. The classification results showed that the MLPNN employing delta-bar-delta training algorithm was effective to detect the ophthalmic arterial Doppler signals with Behcet disease.

Adolescent↗

PubMed1996

Prediction of arterial "burst" activity and transitions between chaotic attractors with a multilayer perceptron optimized by a new stopping criterion.

We have explored the potential of an artificial neural network to capture the dynamics of chaotic temporal fluctuations in arterial pressure and flow. Model generated signals that simulate this ubiquitous physiological phenomenon in both form and complexity were used to train a Multilayer Perceptron (MLP) after first locating the optimum time delay to unfold the attractor governing the dynamics. Prediction horizons were maximized with a new stopping criterion capable of continuously tracking the trajectories of the model system. Single-step predictions were consistently good throughout the study. Long-term predictions obtained by using the MLP as a signal generator were very successful when the number of hidden nodes was carefully chosen. Moreover, short- and long-term predictions could also be obtained even when the dynamics was nonstationary.

Algorithms↗

PubMed2002

Fast accurate MEG source localization using a multilayer perceptron trained with real brain noise.

Iterative gradient methods such as Levenberg-Marquardt (LM) are in widespread use for source localization from electroencephalographic (EEG) and magnetoencephalographic (MEG) signals. Unfortunately, LM depends sensitively on the initial guess, necessitating repeated runs. This, combined with LM's high per-step cost, makes its computational burden quite high. To reduce this burden, we trained a multilayer perceptron (MLP) as a real-time localizer. We used an analytical model of quasistatic electromagnetic propagation through a spherical head to map randomly chosen dipoles to sensor activities according to the sensor geometry of a 4D Neuroimaging Neuromag-122 MEG system, and trained a MLP to invert this mapping in the absence of noise or in the presence of various sorts of noise such as white Gaussian noise, correlated noise, or real brain noise. A MLP structure was chosen to trade off computation and accuracy. This MLP was trained four times, with each type of noise. We measured the effects of initial guesses on LM performance, which motivated a hybrid MLP-start-LM method, in which the trained MLP initializes LM. We also compared the localization performance of LM, MLPs, and hybrid MLP-start-LMs for realistic brain signals. Trained MLPs are much faster than other methods, while the hybrid MLP-start-LMs are faster and more accurate than fixed-4-start-LM. In particular, the hybrid MLP-start-LM initialized by a MLP trained with the real brain noise dataset is 60 times faster and is comparable in accuracy to random-20-start-LM, and this hybrid system (localization error: 0.28 cm, computation time: 36 ms) shows almost as good performance as optimal-1-start-LM (localization error: 0.23 cm, computation time: 22 ms), which initializes LM with the correct dipole location. MLPs trained with noise perform better than the MLP trained without noise, and the MLP trained with real brain noise is almost as good an initial guesser for LM as the correct dipole location.

Brain↗

PubMed2005

Uncertainty of data, fuzzy membership functions, and multilayer perceptrons.

Probability that a crisp logical rule applied to imprecise input data is true may be computed using fuzzy membership function (MF). All reasonable assumptions about input uncertainty distributions lead to MFs of sigmoidal shape. Convolution of several inputs with uniform uncertainty leads to bell-shaped Gaussian-like uncertainty functions. Relations between input uncertainties and fuzzy rules are systematically explored and several new types of MFs discovered. Multilayered perceptron (MLP) networks are shown to be a particular implementation of hierarchical sets of fuzzy threshold logic rules based on sigmoidal MFs. They are equivalent to crisp logical networks applied to input data with uncertainty. Leaving fuzziness on the input side makes the networks or the rule systems easier to understand. Practical applications of these ideas are presented for analysis of questionnaire data and gene expression data.

Algorithms↗