The evolving tree-analysis and applications

In this paper, we enhance and analyze the Evolving Tree (ETree) data analysis algorithm. The suggested improvements aim to make the system perform better while still maintaining the simple nature of the basic algorithm. We also examine the system's behavior with many different kinds of tests, measurements and visualizations. We compare the ETree's performance against classical data analysis methods and very similar modern systems. We find that the ETree is a suitable method for unsupervised analysis of huge data sets.     

Nonlinear dynamical factor analysis for state change detection

Changes in a dynamical process are often detected by monitoring selected indicators directly obtained from the process observations, such as the mean values or variances. Standard change detection algorithms such as the Shewhart control charts or the cumulative sum (CUSUM) algorithm are often based on such first- and second-order statistics. Much better results can be obtained if the dynamical process is properly modeled, for example by a nonlinear state-space model, and then the accuracy of the model is monitored over time. The success of the latter approach depends largely on the quality of the model. In practical applications like industrial processes, the state variables, dynamics, and observation mapping are rarely known accurately. Learning from data must be used; however, methods for the simultaneous estimation of the state and the unknown nonlinear mappings are very limited. We use a novel method of learning a nonlinear state-space model, the nonlinear dynamical factor analysis (NDFA) algorithm. It takes a set of multivariate observations over time and fits blindly a generative dynamical latent variable model, resembling nonlinear independent component analysis. We compare the performance of the model in process change detection to various traditional methods. It is shown that NDFA outperforms the classical methods by a wide margin in a variety of cases where the underlying process dynamics changes.     

Classification Based on Hybridization of Parametric and Nonparametric Classifiers

Parametric methods of classification assume specific parametric models for competing population densities (e.g., Gaussian population densities can lead to linear and quadratic discriminant analysis) and they work well when these model assumptions are valid. Violation in one or more of these parametric model assumptions often leads to a poor classifier. On the other hand, nonparametric classifiers (e.g., nearest-neighbor and kernel-based classifiers) are more flexible and free from parametric model assumptions. But, the statistical instability of these classifiers may lead to poor performance when we have small numbers of training sample observations. Nonparametric methods, however, do not use any parametric structure of population densities. Therefore, even when one has some additional information about population densities, that important information is not used to modify the nonparametric classification rule. This paper makes an attempt to overcome these limitations of parametric and nonparametric approaches and combines their strengths to develop some hybrid classification methods. We use some simulated examples and benchmark data sets to examine the performance of these hybrid discriminant analysis tools. Asymptotic results on their misclassification rates have been derived under appropriate regularity conditions.     

Tele-Existence Techniques of Heavy Work Vehicles

The problem of finding a feasible and technically simple tele-existence system for teleoperation of heavy work vehicles is considered with the aid of two experimental test series. Several test persons of different ages and experience were used in experiments in which a basic audio/video tele-existence system with head-tracking was modified into different versions and tested. Conclusions indicate that details of the tasks and work environment have an important role in finding the optimal solution.     

Quadratic nonnegative matrix factorization

In Nonnegative Matrix Factorization (NMF), a nonnegative matrix is approximated by a product of lower-rank factorizing matrices. Most NMF methods assume that each factorizing matrix appears only once in the approximation, thus the approximation is linear in the factorizing matrices. We present a new class of approximative NMF methods, called Quadratic Nonnegative Matrix Factorization (QNMF), where some factorizing matrices occur twice in the approximation. We demonstrate QNMF solutions to four potential pattern recognition problems in graph partitioning, two-way clustering, estimating hidden Markov chains, and graph matching. We derive multiplicative algorithms that monotonically decrease the approximation error under a variety of measures. We also present extensions in which one of the factorizing matrices is constrained to be orthogonal or stochastic. Empirical studies show that for certain application scenarios, QNMF is more advantageous than other existing nonnegative matrix factorization methods.     

Seasonality in Estonian Traditional Landscape: The Example of Large Village Swings

Like every nation, Estonians have their own indigenous customs and habits. Village swings and the place of these swings in the landscape are focused upon as an example of such Estonian customs, and changes in landscape as related to seasonality and liminality are discussed. The word ‘swing’ is used to denote a large construction (traditionally made of wood) that is able to carry and swing at least two people. Village swings (for public use) are usually located in the middle of a village and the site is commonly used by young people as a place to meet and have a good time. Swinging takes place mostly in the spring and summer, forming one of the many seasonal activities that make up the Estonian traditional calendar. The seasonal break in swinging activity contributes to the eagerness with which swinging is resumed when spring returns, so seasonality creates frames, with the most valued time being the spring. With swinging, the spring has been celebrated as a very valuable and long-awaited season. Although the religious background for swinging has been forgotten, the place is still special and is visited mostly on certain festive occasions. As there is no comparable alternative to swing sites as a socializing place in the village, they have persisted through the centuries. The importance of preserving such unique seasonal places in Estonia's social landscape is demonstrated.     

Stable SOI micromachined electrostatic AC voltage reference

We have designed and manufactured a micromachined moving plate capacitor to be used as an AC voltage reference in electrical metrology. The reference is based on the characteristic AC current–voltage curve of the component having a maximum, the value of which ideally depends only on the geometry of the component and material properties of single crystalline silicon. The electrode surface stability is essential in this application and hence a new fabrication process has been developed to metallize both surfaces of an electrostatically actuated micromachined structure. The stability of the AC reference voltage at a frequency of 100 kHz and an RMS voltage value 6.4 V was measured to be ±60 ppm over 14 h.     

Characterization of diffraction gratings in a rigorous domain with optical scatterometry: hierarchical neural-network model

Characterization of microstructures with features from submicrometers to hundreds of micrometers requires versatile methods. Profilometry and optical microscopy cannot cope with submicrometer features, and atomic-force microscopy, scanning-electron microscopy, and near-field microscopy are inherently slow, off-line methods. In optical scatterometry, the laser light scattered from a sample is measured and the sample profile is subsequently characterized. We propose the use of a two-stage model based on neural networks: rough categorization followed by refinement, thus reducing the need for prior information on the sample. We simulate the method for a submicrometer diffraction grating characterized by five parameters. It is shown that intensity measurements of few diffraction orders by use only of one wavelength are enough to yield rms errors of less than 2 nm for the parameters (approximately 2-3% of the optimal values of the parameters).     

Interference of S-alkyl derivatives of glutathione with brain ionotropic glutamate receptors

The effects of glutathione, glutathione sulfonate and S-alkyl derivatives of glutathione on the binding of glutamate and selective ligands of ionotropic N-methyl-D-aspartate (NMDA) and non-NMDA receptors were studied with mouse synaptic membranes. The effects of glutathione and its analogues on 45Ca2+ influx were also estimated in cultured rat cerebellar granule cells. Reduced and oxidized glutathione, glutathione sulfonate, S-methyl-, -ethyl-, -propyl-, -butyl- and -pentylglutathione inhibited the Na+-independent binding of L-[3H]glutamate. They strongly inhibited also the binding of (S)-2-amino-3-hydroxy-5-[3H]methyl-4-isoxazolepropionate [3H]AMPA (IC50 values: 0.8-15.9 microM). S-Alkylation of glutathione rendered the derivatives unable to inhibit [3H]kainate binding. The NMDA-sensitive binding of L-[3H]glutamate and the binding of 3-[(R)-2-carboxypiperazin-4-yl][1,2-(3)H]propyl-1-phosphonate ([3H]CPP, a competitive antagonist at NMDA sites) were inhibited by the peptides at micromolar concentrations. The strychnine-insensitive binding of the NMDA coagonist [3H]glycine was attenuated only by oxidized glutathione and glutathione sulfonate. All peptides slightly enhanced the use-dependent binding of [3H]dizocilpine (MK-801) to the NMDA-gated ionophores. This effect was additive with the effect of glycine but not with that of saturating concentrations of glutamate or glutamate plus glycine. The glutamate- and NMDA-evoked influx of 45Ca2+ into cerebellar granule cells was inhibited by the S-alkyl derivatives of glutathione. We conclude that besides glutathione the endogenous S-methylglutathione and glutathione sulfonate and the synthetic S-alkyl derivatives of glutathione act as ligands of the AMPA and NMDA receptors. In the NMDA receptor-ionophore these glutathione analogues bind preferably to the glutamate recognition site via their gamma-glutamyl moieties.     

Blind separation of positive sources by globally convergent gradient search

The instantaneous noise-free linear mixing model in independent component analysis is largely a solved problem under the usual assumption of independent nongaussian sources and full column rank mixing matrix. However, with some prior information on the sources, like positivity, new analysis and perhaps simplified solution methods may yet become possible. In this letter, we consider the task of independent component analysis when the independent sources are known to be nonnegative and well grounded, which means that they have a nonzero pdf in the region of zero. It can be shown that in this case, the solution method is basically very simple: an orthogonal rotation of the whitened observation vector into nonnegative outputs will give a positive permutation of the original sources. We propose a cost function whose minimum coincides with nonnegativity and derive the gradient algorithm under the whitening constraint, under which the separating matrix is orthogonal. We further prove that in the Stiefel manifold of orthogonal matrices, the cost function is a Lyapunov function for the matrix gradient flow, implying global convergence. Thus, this algorithm is guaranteed to find the nonnegative well-grounded independent sources. The analysis is complemented by a numerical simulation, which illustrates the algorithm.