PubMed Health⌕ Search

Biomedical subjects

P Tavan

Publications and source records attributed to P Tavan.

11 recordsLinked to original sources
PubMed2000

Generalized radial basis function networks for classification and novelty detection: self-organization of optimal Bayesian decision.

By adding reverse connections from the output layer to the central layer it is shown how a generalized radial basis functions (GRBF) network can self-organize to form a Bayesian classifier, which is also capable of novelty detection. For this purpose, three stochastic sequential learning rules are introduced from biological considerations which pertain to the centers, the shapes, and the widths of the receptive fields of the neurons and allow ajoint optimization of all network parameters. The rules are shown to generate maximum-likelihood estimates of the class-conditional probability density functions of labeled data in terms of multivariate normal mixtures. Upon combination with a hierarchy of deterministic annealing procedures, which implement a multiple-scale approach, the learning process can avoid the convergence problems hampering conventional expectation-maximization algorithms. Using an example from the field of speech recognition, the stages of the learning process and the capabilities of the self-organizing GRBF classifier are illustrated.

Artificial Intelligence↗
PubMed1996

Ligand binding: molecular mechanics calculation of the streptavidin-biotin rupture force.

The force required to rupture the streptavidin-biotin complex was calculated here by computer simulations. The computed force agrees well with that obtained by recent single molecule atomic force microscope experiments. These simulations suggest a detailed multiple-pathway rupture mechanism involving five major unbinding steps. Binding forces and specificity are attributed to a hydrogen bond network between the biotin ligand and residues within the binding pocket of streptavidin. During rupture, additional water bridges substantially enhance the stability of the complex and even dominate the binding interactions. In contrast, steric restraints do not appear to contribute to the binding forces, although conformational motions were observed.

Bacterial Proteins↗
PubMed1994

An associative memory that can form hypotheses: a phase-coded neural network.

Nonlinear associative memories as realized, e.g., by Hopfield nets are characterized by attractor-type dynamics. When fed with a starting pattern, they converge to exactly one of the stored patterns which is supposed to be most similar. These systems cannot render hypotheses of classification, i.e., render several possible answers to a given classification problem. Inspired by von der Malsburg's correlation theory of brain function, we extend conventional neural network architectures by introducing additional dynamical variables. Assuming an oscillatory time structure of neural firing, i.e., the existence of neural clocks, we assign a so-called phase to each formal neuron. The phases explicitly describe detailed correlations of neural activities neglected in conventional neural network architectures. Implementing this extension into a simple self-organizing network based on a feature map, we present an associative memory that actually is capable of forming hypotheses of classification.

Animals↗
PubMed1991

[Bone histomorphometric study in involuted fractured osteoporosis treated with 1-ethane-1-hydroxybiphosphonate (etidronate) during one year].

The efficacy of the biphosphonate etidronate has recently been demonstrated versus the vertebral fracture rate in fractured involuted osteoporosis in the literature. However, it would appear that the increase in bone mass (measured from the calcium density) may alone account for these results. The authors undertook a histomorphometrie study in 20 patients with a group mean age of 55 years presenting recent vertebral fracture in the context of osteoporosis, in order to assess the cell changes which may affect bone quality. This factor remains the only one which can account for the reduced number of fractures. The patients received treatment with phosphorus (1,500 mg/d) for 3 days followed by etidronate (400 mg/d) for 14 days every 90 days. A permanent daily intake of 50 mg of calcium and 400 IU of vitamin D was also administered. Each patient underwent bone biopsy of the iliac wing before and after one year of treatment (4 cycles). No change in the bone mass or architectural parameters was observed. However, general slowing of the bone remodeling was found, affecting the natality and activity of the osteoforming and osteoresorbing cells. However, this remodeling which is traditionally uncoupled in osteoporosis was once more coupled. This results in a slowing of bone loss, ageing of the bone present and no change in the architecture. Thus, the diphosphonates appear to have a beneficial effect on the quality of bone rather than on its quantity. This represents a novel approach to the treatment of osteoporosis.

Aged↗
PubMed1990

Self-organization of associative memory and pattern classification: recurrent signal processing on topological feature maps.

We extend the neural concepts of topological feature maps towards self-organization of auto-associative memory and hierarchical pattern classification. As is well-known, topological maps for statistical data sets store information on the associated probability densities. To extract that information we introduce a recurrent dynamics of signal processing. We show that the dynamics converts a topological map into an auto-associative memory for real-valued feature vectors which is capable to perform a cluster analysis. The neural network scheme thus developed represents a generalization of non-linear matrix-type associative memories. The results naturally lead to the concept of a feature atlas and an associated scheme of self-organized, hierarchical pattern classification.

Algorithms↗
PubMed1985

A model for the lateral diffusion of "stiff" chains in a lipid bilayer.

We present random walk models for the diffusive motion of lipid probe molecules in a lipid bilayer and calculate the diffusion constants for probes spanning the entire bilayer and for probes extending through one lipid layer only. The "stiffness" of such molecules can explain the observed value of 2/3 for the ratio of these diffusion constants.

Diffusion↗