Artificial neural network learning of nonstationary behavior in time series

We refine and complement a previously-proposed artificial neural network method for learning hidden signals forcing nonstationary behavior in time series. The method adds an extra input unit to the network and feeds it with the proposed profile for the unknown perturbing signal. The correct time evolution of this new input parameter is learned simultaneously with the intrinsic stationary dynamics underlying the series, which is accomplished by minimizing a suitably-defined error function for the training process. We incorporate here the use of validation data, held out from the training set, to accurately determine the optimal value of a hyperparameter required by the method. Furthermore, we evaluate this algorithm in a controlled situation and show that it outperforms other existing methods in the literature. Finally, we discuss a preliminary application to the real-world sunspot time series and link the obtained hidden perturbing signal to the secular evolution of the solar magnetic field.     

Facet shapes and thermo-stabilities of H<Subscript>2</Subscript>SO<Subscript>4</Subscript>•HNO<Subscript>3</Subscript> hydrates involved in polar stratospheric clouds

The nucleation, ice crystal shapes and thermodynamic stability of polar stratospheric clouds particles are interesting concerns owing to their implication in the ozone layer destruction. Some of these particles are formed by conformers of H₂O, HNO₃, and H₂SO₄. We carried out calculations using density functional theory (DFT) to obtain optimized structures. Several stable trimers are achieved —divided in two groups, one with HNO₃ moiety, second with H₂SO₄ moiety— after pre-optimization at B3LYP/6-31G and subsequently optimization at B3LYP/aug-cc-pVTZ level of theory. For both most stable conformers five H₂O molecules are added to their optimized trimers to calculate hydrated geometries. The OH stretching harmonic frequencies are provided for all aggregates. The zero-point energy correction (ZEPC), relative electronic energies (?E), relative reaction Gibbs free energies ?(?G)_k-relative, and cooling constant (K _cooling ) are reported at three temperatures: 188 K, 195 K, and 210 K. Shapes given in our calculations are compared with various experimental shapes as well as comparisons with their thermo-stabilities.

Open image in new window

Graphical Abstract Facet shapes and thermo-stabilities of H₂SO₄?HNO₃ hydrates involved in polar stratospheric clouds. IR spectrum of WNS-1+5W structure and its circular facet

     

Confocal reader for biochip screening and fluorescence microscopy

We developed a fluorescence reader for the sensitive detection of surface-generated fluorescence. The system is applicable for high resolution imaging as well as for the readout of large biochips. The surface of a microscope coverslip is scanned with a laser beam focused to a waist diameter of 500 nm (FWHM) by means of a single aspheric lens. Scanning large areas with a focused beam usually evokes the need of automatic control elements to adjust the laser spot to the designated position at the surface. Due to the special design of the reader, the focus keeps at the plane of the surface even when scanning large areas, obviating the requirement of any real time control. Thus the instrument is straightforward and inexpensive. Nevertheless it features a high sensitivity and high optical resolution. The versatility of the instrument is demonstrated by imaging cells and reading out a DNA-chip. The excellent sensitivity is shown by detecting single fluorescently labeled antibodies.     

Nature’s Palette: Characterization of Shared Pigments in Colorful Avian and Mollusk Shells

Pigment-based coloration is a common trait found in a variety of organisms across the tree of life. For example, calcareous avian eggs are natural structures that vary greatly in color, yet just a handful of tetrapyrrole pigment compounds are responsible for generating this myriad of colors. To fully understand the diversity and constraints shaping nature’s palette, it is imperative to characterize the similarities and differences in the types of compounds involved in color production across diverse lineages. Pigment composition was investigated in eggshells of eleven paleognath bird taxa, covering several extinct and extant lineages, and shells of four extant species of mollusks. Birds and mollusks are two distantly related, calcareous shell-building groups, thus characterization of pigments in their calcareous structures would provide insights to whether similar compounds are found in different phyla (Chordata and Mollusca). An ethylenediaminetetraacetic acid (EDTA) extraction protocol was used to analyze the presence and concentration of biliverdin and protoporphyrin, two known and ubiquitous tetrapyrrole avian eggshell pigments, in all avian and molluscan samples. Biliverdin was solely detected in birds, including the colorful eggshells of four tinamou species. In contrast, protoporphyrin was detected in both the eggshells of several avian species and in the shells of all mollusks. These findings support previous hypotheses about the ubiquitous deposition of tetrapyrroles in the eggshells of various bird lineages and provide evidence for its presence also across distantly related animal taxa.     

Psychotic symptoms are associated with physical health problems independently of a mental disorder diagnosis: results from the WHO World Health Survey

This study explored whether physical health problems are related to psychotic symptoms independently of a mental disorder diagnosis. A total of 224,254 subjects recruited for the World Health Organization World Health Survey were subdivided into those with both a lifetime diagnosis of psychosis and at least one psychotic symptom in the 12 months prior to the evaluation, those with at least one psychotic symptom in the past 12 months but no lifetime diagnosis of psychosis, and those without psychotic symptoms in the past 12 months and without a lifetime diagnosis of psychosis. The three groups were compared for the presence of medical conditions, health problems, and access to health care. Medical conditions and health problems (angina, asthma, arthritis, tuberculosis, vision or hearing problems, mouth/teeth problems, alcohol consumption, smoking, and accidents), medication consumption, and hospital admissions (but not regular health care visits) were more frequent in individuals with psychotic symptoms but no psychosis diagnosis, compared to those with no symptoms and no diagnosis. The number of medical conditions increased with the number of psychotic symptoms. Given the sample analyzed, this trend seems to be independent from the socio‐economic development of the country or the specific health care system.     

A late-stopping method for optimal aggregation of neural networks

Ensembles of artificial neural networks have been used in the last years as classification/regression machines, showing improved generalization capabilities that outperform those of single networks. However, it has been recognized that for aggregation to be effective the individual networks must be as accurate and diverse as possible. An important problem is, then, how to tune the aggregate members in order to have an optimal compromise between these two conflicting conditions. We propose here a simple method for constructing regression/classification ensembles of neural networks that leads to overtrained aggregate members with an adequate balance between accuracy and diversity. The algorithm is favorably tested against other methods recently proposed in the literature, producing an improvement in performance on the standard statistical databases used as benchmarks. In addition, and as a concrete application, we apply our method to the sunspot time series and predict the remainder of the current cycle 23 of solar activity.