Siamese multi-layer perceptrons for dimensionality reduction and face identification

This paper presents a framework using siamese Multi-layer Perceptrons (MLP) for supervised dimensionality reduction and face identification. Compared with the classical MLP that trains on fully labeled data, the siamese MLP learns on side information only, i.e., how similar of data examples are to each other. In this study, we compare it with the classical MLP on the problem of face identification. Experimental results on the Extended Yale B database demonstrate that the siamese MLP training with side information achieves comparable classification performance with the classical MLP training on fully labeled data. Besides, while the classical MLP fixes the dimension of the output space, the siamese MLP allows flexible output dimension, hence we also apply the siamese MLP for visualization of the dimensionality reduction to the 2-d and 3-d spaces.     

Scheduling the professional soccer leagues of Austria and Germany

Generating a regular season schedule is a demanding task for any sports league. In Europe, the creation of a suitable schedule for every national top soccer league not only has to address numerous conflicting inner-league requirements and preferences. Additionally, the games of the European Cup matches (Champions League, UEFA Cup, National Cup Winners) have to be taken into account. In this paper we consider the case of Austria and Germany, that is the planning problem the “Deutsche Fußball-Bund” (DFB) and the “Österreichische Fußball-Bund” (ÖFB) are confronted with. For both leagues we develop models and algorithms which yield reasonable schedules quickly. The models borrow their expressive power from so-called partially renewable resources. Our approach generates schedules which have been accepted for play once by the DFB and six times by the ÖFB.     

Resolution of singularities for multi-loop integrals

We report on a program for the numerical evaluation of divergent multi-loop integrals. The program is based on iterated sector decomposition. We improve the original algorithm of Binoth and Heinrich such that the program is guaranteed to terminate. The program can be used to compute numerically the Laurent expansion of divergent multi-loop integrals regulated by dimensional regularisation. The symbolic and the numerical steps of the algorithm are combined into one program.

Program summary

Program title: sector_decompositionCatalogue identifier: AEAG_v1_0Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AEAG_v1_0.htmlProgram obtainable from: CPC Program Library, Queen's University, Belfast, N. IrelandLicensing provisions: Standard CPC licence, http://cpc.cs.qub.ac.uk/licence/licence.htmlNo. of lines in distributed program, including test data, etc.: 47 506No. of bytes in distributed program, including test data, etc.: 328 485Distribution format: tar.gzProgramming language: C++Computer: allOperating system: UnixRAM: Depending on the complexity of the problemClassification: 4.4External routines: GiNaC, available from http://www.ginac.de, GNU scientific library, available from http://www.gnu.org/software/gslNature of problem: Computation of divergent multi-loop integrals.Solution method: Sector decomposition.Restrictions: Only limited by the available memory and CPU time.Running time: Depending on the complexity of the problem.     

Histone Methylation Marks on Circulating Nucleosomes as Novel Blood-Based Biomarker in Colorectal Cancer

Circulating nucleic acids (CNAs) are under investigation as a liquid biopsy in cancer as potential non-invasive biomarkers, as stable structure in circulation nucleosomes could be valuable sources for detection of cancer-specific alterations in histone modifications. Our interest is in histone methylation marks with a focus on colorectal cancer, one of the leading cancers respective the incidence and mortality. Our previous work included the analysis of trimethylations of lysine 9 on histone 3 (H3K9me3) and of lysine 20 on histone 4 (H4K20me3) by chromatin immuno- precipitation-related PCR in circulating nucleosomes. Here we asked whether global immunologic measurement of histone marks in circulation could be a suitable approach to show their potential as biomarkers. In addition to H3K9me3 and H4K20me3 we also measured H3K27me3 in plasma samples from CRC patients (n = 63) and cancer free individuals (n = 40) by ELISA-based methylation assays. Our results show that of three marks, the amounts of H3K27me3 (p = 0.04) and H4K20me3 (p < 0.001) were significantly lower in CRC patients than in healthy controls. For H3K9me3 similar amounts were measured in both groups. Areas under the curve (AUC) in receiver operating characteristic (ROC) curves indicating the power of CRC detection were 0.620 for H3K27me3, 0.715 for H4K20me3 and 0.769 for the combination of both markers. In conclusion, findings of this preliminary study reveal the potential of blood-based detection of CRC by quantification of histone methylation marks and the additive effect of the marker combination.     

The Global Biogeochemical Silicon Cycle

Silicon is one of the most important elements in the current age of the anthropocene. It has numerous industrial applications, and supports a high-tech multi-billion Euro industry. Silicon has a fascinating biological and geological cycle, interacting with other globally important biogeochemical cycles. In this review, we bring together both biological and geological aspects of the silicon cycle to provide a general, comprehensive review of the cycling of silicon in the environment. We hope this review will provide inspiration for researchers to study this fascinating element, as well as providing a background environmental context to those interested in silicon.     

Synthesis of 2‐[18F]Fluoro‐2‐deoxyisosorbide 5‐mononitrate and Assessment of Its in vivo Biodistribution as Determined by Dynamic Positron Emission Tomography (PET)

Herein we disclose the synthesis of 2‐fluoro‐2‐deoxyisosorbide 5‐mononitrate (2F‐IS‐5MN), a fluorinated analogue of the commonly prescribed vasodilator isosorbide 5‐mononitrate (IS‐5MN). X‐ray structural data for IS‐5MN and its C2‐epimeric congener IM‐5MN are presented together with structural data for 2F‐IS‐5MN. Radioisotope labeling of 2F‐IS‐5MN has, for the first time, allowed observation of the in vivo biodistribution of this organic nitrate by means of dynamic positron emission tomography (PET) in wild‐type mice.