Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Hidden variables”

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 217 records · Page 12Linked to original sources
PubMedDate not supplied

[Prediction of binding affinities of protein-ligand complexes using nonlinear models].

A network model for prediction of the free energy changes in protein-ligand complexes has been developed. The 150 complexes of different nature were used as a training set. The computational physics-chemical parameters of these complexes were used as independent variables. Both classical models of multiple linear regression and several network models with one hidden layer were created and the best was chosen. Significant improvement was shown for network model prediction quality in comparison with classical model of multiple linear regression (R2 on training--0.81 and 0.54; R2 on "leave-one-out" procedure--0.74 and 0.52 respectively).

Ligands↗
PubMed2002

Application of hidden Markov models on residuals: an example using Canadian traffic accident data.

Laverty, Kelly, Rotton, and Flynn conducted a regression analysis in 1992 on 9 years of automobile accidents in Saskatchewan (a total of 200,545 accidents) to find a small linear trend, season effects, holiday, and day of the week effects. The application of a hidden Markov model to the residuals of this analysis uncovered two states which are likely to be related to the weather. These states can be described as low volatility' and 'high volatility'. The 'low volatility' state involves low variability compared to the 'high volatility' state (occurring during the colder months) during which the largest numbers of accidents occur. It is suggested that hidden Markov models are a useful method for uncovering hidden, underlying states in social science and health-related data.

Accidents, Traffic↗
PubMed1999

Sonographic prediction of malignancy in adnexal masses using an artificial neural network.

OBJECTIVE: To generate a neural network algorithm which computes a probability of malignancy score for pre-operative discrimination between malignant and benign adnexal tumours. DESIGN: A retrospective analysis of previously collected data. Information from 75% of the study group was used to train an artificial neural network and the remainder was used for validation. SETTING: The Gynaecological Ultrasound Research Unit at King's College Hospital, London. POPULATION: Sixty-seven women with known adnexal mass who had been examined using transvaginal B-mode ultrasonography and colour Doppler imaging with pulse spectral analysis immediately before surgery. The excised masses were classified histologically as benign (n = 52) or malignant (n = 15), of which three were borderline. METHODS: The variables that were put into the artificial neural network were: age, menopausal status, maximum tumour diameter, tumour volume, locularity, the presence of papillary projections, the presence of random echogenicity, the presence of analysable blood flow velocity waveforms, the peak systolic velocity, time-averaged maximum velocity, the pulsatility index, and resistance index. Histological classification, categorised as benign or malignant, was the output result. RESULTS: A variant of the back propagation method was selected to train the network. The overall architecture of the network with the best performance contained an input layer with four variables (age, time-averaged maximum velocity, papillary projection score and maximum tumour diameter), a hidden layer with three units and an output layer with one. The sensitivity and specificity at the optimum diagnostic decision value for the artificial neural network output (0.45) were 100% (95% CI 78.2%-100%) and 98.1% (95% CI 89.5%-100%), respectively. These values were significantly better than those obtained from the independent use of the resistance index, pulsatility index, time-averaged maximum velocity or peak systolic velocity at their optimum decision values (P < 0.01). CONCLUSION: Artificial neural networks may be used on clinical and ultrasound derived end-points to accurately predict ovarian malignancy. There is a need for a prospective evaluation of this technique using a larger number of patients.

Adnexal Diseases↗
PubMed1999

Classification of dolphin echolocation clicks by energy and frequency distributions.

Dolphins demonstrate an adaptive control over echolocation click production, but little is known of the manner or degree with which control is exercised. Echolocation clicks (N approximately 30,000) were collected from an Atlantic bottlenose dolphin (Tursiops truncatus) performing object discrimination tasks in order to investigate differential click production. Seven categories of clicks were identified using the spectral conformation and relative position of -3 and -10 dB peaks. A counterpropagation network utilizing 16 inputs, 50 hidden units, and 8 output units was trained to classify clicks using the same spectral variables. The network classified novel clicks with 92% success. Additional echolocation clicks (N > 24,000) from two other dolphins were submitted to the network for classification. Classified echolocation clicks were analyzed for animal specific differences, changes in predominant click type within click trains, and task-related specificity. Differences in animal and task performance may influence click type and click train length.

Animals↗
PubMed2001

The dual-task SRT procedure: fine-tuning the timing.

In the standard sequential reaction time study, subjects are presented with a repeating sequence of targets to which they must respond as rapidly as possible. With practice reaction times decrease, suggesting a learned ability to exploit the repeating patterns in the display. In a common variation, a second task is interposed, typically a tone-counting task in which subjects must keep track of the frequency with which particular tones occur. In the canonical experiment, these tones appear at varying times in the interval between the subjects' response to a target and the next target. Data are presented that show that this latter variable (the RSOA, response-secondary stimulus onset asynchrony) actually plays an important but previously hidden role in these experiments. A model based on an extension of the notion of the psychological refractory period is introduced to explain these findings.

Adult↗
PubMed1992

Anti-histone autoantibodies recognize centromeric heterochromatin in metaphase chromosomes and hidden epitopes in interphase cells.

IgM autoantibodies from a subset of patients with undifferentiated rheumatic disease syndromes stained mouse kidney nuclei with a distinctive variable large-speckled (VLS) indirect immunofluorescence (IIF) pattern. However, these antibodies did not stain nuclei of tissue culture cells prepared with conventional fixation. These sera were shown to react with histone H3 by an enzyme-linked immunosorbent assay (ELISA), and adsorption with H3 reduced or eliminated the IIF reaction. Sera yielding a VLS-IIF pattern reacted in ELISA with all three H3 variants as well as the native (H3-H4)2 tetramer, but the reactive determinants were unavailable for binding when chromatin was the substrate. By IIF assay, the epitopes were exposed after treatment of tissue culture cells with 0.5 M NaCl, and were removed by 1.5 M NaCl. These sera also stained the centromeric region of metaphase chromosomes. These observations suggest that the VLS-IIF pattern is due to antibodies that recognize epitopes on constitutive heterochromatin near the centromere. The epitopes are exposed in differentiated cells but hidden in dividing cells. Histone in heterochromatin, or CENP-A, a histone-like protein in the centromere with a sequence similarity to histone H3, are candidates for the target antigen.

Autoantibodies↗
PubMed1996

A Numerical Implementation of Kolmogorov's Superpositions.

Hecht-Nielsen proposed a feedforward neural network based on Kolmogorov's superpositionsf(x(i), em leader,x(n))= summation operator q=02piPhi(q)(y(q)) that apply to all real valued continuous functions f(x(1), em leader, x(n)) defined on a Euclidean unit cube of dimension n >/= 2. This network has a hidden layer that is independent of f and that transforms the n-tuples (x(1), em leader, x(n)) into the 2n + 1 variables y(q), and an output layer in which f is computed. Kůrková has shown that such a network has an approximate implementation with arbitrary activation functions of sigmoidal type. Actual implementation is, however, impeded by the lack of numerical algorithms for the hidden layer which contain continuous functions of the formy(q)=Sigma(n)(p=1)alpha(p)psi(x(p)+qa) with constants a and alpha(p). This paper gives an explicit numerical implementation of the hidden layer that also enables the implementation of the output layer. Copyright 1996 Elsevier Science Ltd

Journal Article↗
PubMed1979

Intracellular synaptic potentials of primate motor cortex neurons during voluntary movement.

An intracellular recording technique was applied to the precentral motor cortex of the unanesthetized, chronically behaving monkey. Postsynaptic potentials, responsible for an initiation of the voluntary movement, were recorded. In total, 22 pyramidal tract neurons (PTNs) and 40 non-pyramidal tract neurons (non-PTNs) were successfully penetrated in 5 monkeys while the monkey was performing a flexion-extension wrist movement after a visual cue (reaction time, 200--350 msec). The neurons showed a negative membrane potential shift of at least 30 mV for more than 30 sec. A slowly rising PSP appeared 80--180 msec after the visual cue, and was 70--180 msec prior to an onset of the movement. Spike activities were superimposed upon this slow PSP with 20--80 msec rise time and 2--6 mV depolarization (8 PTNs and 6 non-PTNs). Since these depolarizations were variable in magnitude and latency, these were considered to be summated potentials of small EPSPs and hidden IPSPs. Membrane resistance was measured from an IR drop by a hyperpolarizing current (1.2 X 10(-9) A) passed through a recording electrode, and was 3.5 +/- 1.7 Momega in 18 PTNs and 4.5 +/- 2.5 Momega in 28 non-PTNs. There was a linear relationship in PTNs between membrane resistance and antidromic latency from the pontine pyramid. Because of the time course of PSPs, their possible dendritic origins were discussed.

Animals↗
PubMed2004

Multivariate approaches in plant science.

The objective of proteomics is to get an overview of the proteins expressed at a given point in time in a given tissue and to identify the connection to the biochemical status of that tissue. Therefore sample throughput and analysis time are important issues in proteomics. The concept of proteomics is to encircle the identity of proteins of interest. However, the overall relation between proteins must also be explained. Classical proteomics consist of separation and characterization, based on two-dimensional electrophoresis, trypsin digestion, mass spectrometry and database searching. Characterization includes labor intensive work in order to manage, handle and analyze data. The field of classical proteomics should therefore be extended to also include handling of large datasets in an objective way. The separation obtained by two-dimensional electrophoresis and mass spectrometry gives rise to huge amount of data. We present a multivariate approach to the handling of data in proteomics with the advantage that protein patterns can be spotted at an early stage and consequently the proteins selected for sequencing can be selected intelligently. These methods can also be applied to other data generating protein analysis methods like mass spectrometry and near infrared spectroscopy and examples of application to these techniques are also presented. Multivariate data analysis can unravel complicated data structures and may thereby relieve the characterization phase in classical proteomics. Traditionally statistical methods are not suitable for analysis of the huge amounts of data, where the number of variables exceed the number of objects. Multivariate data analysis, on the other hand, may uncover the hidden structures present in these data. This study takes its starting point in the field of classical proteomics and shows how multivariate data analysis can lead to faster ways of finding interesting proteins. Multivariate analysis has shown interesting results as a supplement to classical proteomics and added a new dimension to the field of proteomics.

Algorithms↗
PubMed2004

Quantitative structure-metabolism relationship modeling of metabolic N-dealkylation reaction rates.

It is widely recognized that preclinical drug discovery can be improved via the parallel assessment of bioactivity, absorption, distribution, metabolism, excretion, and toxicity properties of molecules. High-throughput computational methods may enable such assessment at the earliest, least expensive discovery stages, such as during screening compound libraries and the hit-to-lead process. As an attempt to predict drug metabolism and toxicity, we have developed an approach for evaluation of the rate of N-dealkylation mediated by two of the most important human cytochrome P450s (P450), namely CYP3A4 and CYP2D6. We have taken a novel approach by using descriptors generated for the whole molecule, the reaction centroid, and the leaving group, and then applying neural network computations and sensitivity analysis to generate quantitative structure-metabolism relationship models. The quality of these models was assessed by using the cross-validated correlation coefficients of 0.82 for CYP3A4 and 0.79 for CYP2D6 as well as external test molecules for each enzyme. The relative performance of different neural networks was also compared, and modular neural networks with two hidden layers provided the best predictive ability. Functional dependencies between the neural network input and output variables, generalization ability, and limitations of the described approach are also discussed. These models represent an initial approach to predicting the rate of P450-mediated metabolism and may be applied and integrated with other models for P450 binding to produce a systems-based approach for predicting drug metabolism.

Cytochrome P-450 CYP2D6↗
PubMed1999

Feedforward neural network with adaptive reference pattern layer.

A hybrid neural network architecture is investigated for modeling purposes. The proposed hybrid is based on the multilayer perceptron (MLP) network. In addition to the usual hidden layers, the first hidden layer is selected to be an adaptive reference pattern layer. Each unit in this new layer incorporates a reference pattern that is located somewhere in the space spanned by the input variables. The outputs of these units are the component wise-squared differences between the elements of a reference pattern and the inputs. The reference pattern layer has some resemblance to the hidden layer of the radial basis function (RBF) networks. Therefore the proposed design can be regarded as a sort of hybrid of MLP and RBF networks. The presented benchmark experiments show that the proposed hybrid can provide significant advantages over standard MLPs and RBFs in terms of fast and efficient learning, and compact network structure.

Feedback↗
PubMed2004

[Identification of official rhubarb samples based on NIR spectra and neural networks].

Rhubarb is one of the most widely used Chinese medicinal herbs in China. Fast and accurate identification of official and unofficial rhubarb samples is most critical for quality control of Chinese medicine production. In the present paper near-infrared reflectance spectrometry (NIRS) and artificial neural network (ANN) were combined to develop classification models for identifying 52 official and unofficial rhubarb samples. The measured spectra were compressed by wavelet transformation (WT) and then the ANN classification models were trained with the reduced-variables spectral data. The rate of correct classification was over 96%. The effects of neurons in hidden layer and the momentum were also discussed. Owing to its fast and nondestructive properties, NIRS is a promising approach to identifying Chinese medicinal herbs.

Artificial Intelligence↗
PubMedDate not supplied

Examination of microheterogeneity in grass pollen allergens.

The separation of timothy pollen extract by two-dimensional immunoblotting revealed microheterogeneity of the major allergens PhI p I and PhI p V. There was not only a diversity in size, 38 and 32 kDa for PhI p V and 37, 35, and 33 kDa for PhI p I, but also a separation into proteins of identical sizes but different pIs. Since former studies on the protein structure by amino acid analysis and N-terminal microsequencing did not reveal any differences, we examined other possibilities that might cause microheterogeneity. In allergens belonging to the PhI p I group, the variability in pI can be due to the carbohydrate structure and to the fact that charges are hidden in the interior of the protein, as shown by varying concentrations of urea. On the other hand we were not able to detect any such reasons for the existence of PhI p V isoallergens. Thus, we assume that they are stable conformational isomers of the proteins (allomorphism) or that there are only slight variations in the internal sequences of these proteins (polymorphism) causing distinct pIs.

Allergens↗
PubMed1998

Personality comparison between children of hidden Holocaust survivors and American Jewish parents.

The possibility that the experiences of the "hidden" child survivors of the Holocaust (those who survived outside of the concentration camps during the Nazi occupation) had a pathological effect on their offspring was examined by comparing volunteer, matched samples of adult children of "hidden" child survivors of the Holocaust with adult children of nontraumatized U.S.-born Jewish parents on personality variables measured by the Sixteen Personality Factor Questionnaire (Cattell, Eber, & Tatsuoka, 1970). The MANOVA results indicated that there were no differences in the personality characteristics of the two groups.

Adult↗
PubMed1995

Drosophila lactate dehydrogenase. Functional and evolutionary aspects.

The functional characteristics of homogeneously purified LDH were studied in the eight D. melanogaster species subgroup at two different growing temperatures (14 degrees C, 25 degrees C). The Vmax, kcat, Vmax/KeNAD.KmLac and Kcat/KsNAD KmLac values detected at the permissive growing temperature of 25 degrees C, were found to converge with the consensus phylogeny of these species which consists of two (melanogaster, yakuba) complexes. This scheme, also verified by the Micro-Complement Fixation (MC'F) method in another study in our Laboratory, substantiates the connection between the enzyme function and the phylogeny of these species. We propose that the major variation of the Ldh gene in this subgroup has arisen prior to the first species divergence, the final result of which is the eight sibling species. On the other hand, the variable catalytic differentiation observed at the restrictive temperature of 14 degrees C may enrich the species with hidden adaptive possibilities.

Animals↗
PubMed2002

Examination of residuals to Vancouver crisis call data by using hidden Markov models.

Three years of daily crisis calls (N=2,513,478) were examined in a regression analysis with terms for effects for weekday, end and beginning of the month, holiday, seasonal trend, and long-term linear trend. This was followed by an analysis of residuals using hidden Markov models, with three states best fitting the residuals. These states may reflect economic or environmental variables which researchers should clarify in further work.

British Columbia↗
PubMed1995

A circadian perspective on myocardial ischaemia.

The non-uniform occurrence of myocardial ischaemia, with a trough at night and a peak in the morning, is at the base of the concept of a circadian distribution of coronary events. Cardiovascular variables known to influence the occurrence of ischaemia have similar rhythms, and a likely culprit could be hidden among them. Nevertheless, in many cases an identifiable trigger is lacking and the prediction of ischaemia is still elusive.

Autonomic Nervous System↗
PubMed2003

A hierarchical method for finding optimal architecture and weights using evolutionary least square based learning.

In this paper, we present a novel approach of implementing a combination methodology to find appropriate neural network architecture and weights using an evolutionary least square based algorithm (GALS).1 This paper focuses on aspects such as the heuristics of updating weights using an evolutionary least square based algorithm, finding the number of hidden neurons for a two layer feed forward neural network, the stopping criterion for the algorithm and finally some comparisons of the results with other existing methods for searching optimal or near optimal solution in the multidimensional complex search space comprising the architecture and the weight variables. We explain how the weight updating algorithm using evolutionary least square based approach can be combined with the growing architecture model to find the optimum number of hidden neurons. We also discuss the issues of finding a probabilistic solution space as a starting point for the least square method and address the problems involving fitness breaking. We apply the proposed approach to XOR problem, 10 bit odd parity problem and many real-world benchmark data sets such as handwriting data set from CEDAR, breast cancer and heart disease data sets from UCI ML repository. The comparative results based on classification accuracy and the time complexity are discussed.

Algorithms↗