Feature selection by multi-objective optimisation: Application to network anomaly detection by hierarchical self-organising maps

Feature selection is an important and active issue in clustering and classification problems. By choosing an adequate feature subset, a dataset dimensionality reduction is allowed, thus contributing to decreasing the classification computational complexity, and to improving the classifier performance by avoiding redundant or irrelevant features. Although feature selection can be formally defined as an optimisation problem with only one objective, that is, the classification accuracy obtained by using the selected feature subset, in recent years, some multi-objective approaches to this problem have been proposed. These either select features that not only improve the classification accuracy, but also the generalisation capability in case of supervised classifiers, or counterbalance the bias toward lower or higher numbers of features that present some methods used to validate the clustering/classification in case of unsupervised classifiers. The main contribution of this paper is a multi-objective approach for feature selection and its application to an unsupervised clustering procedure based on Growing Hierarchical Self-Organising Maps (GHSOMs) that includes a new method for unit labelling and efficient determination of the winning unit. In the network anomaly detection problem here considered, this multi-objective approach makes it possible not only to differentiate between normal and anomalous traffic but also among different anomalies. The efficiency of our proposals has been evaluated by using the well-known DARPA/NSL-KDD datasets that contain extracted features and labelled attacks from around 2 million connections. The selected feature sets computed in our experiments provide detection rates up to 99.8% with normal traffic and up to 99.6% with anomalous traffic, as well as accuracy values up to 99.12%.

Electrochemical Performance of Hierarchical Porous Carbon Materials Obtained from the Infiltration of Lignin into Zeolite Templates

Hierarchical porous carbon materials prepared by the direct carbonization of lignin/zeolite mixtures and the subsequent basic etching of the inorganic template have been electrochemically characterized in acidic media. These lignin-based templated carbons have interesting surface chemistry features, such as a variety of surface oxygen groups and also pyridone and pyridinic groups, which results in a high capacitance enhancement compared to petroleum-pitch-based carbons obtained by the same procedure. Furthermore, they are easily electro-oxidized in a sulfuric acid electrolyte under positive polarization to produce a large amount of surface oxygen groups that boosts the pseudocapacitance. The lignin-based templated carbons showed a specific capacitance as high as 250 F g−1 at 50 mA g−1, with a capacitance retention of 50 % and volumetric capacitance of 75 F cm−3 at current densities higher than 20 A g−1 thanks to their suitable porous texture. These results indicate the potential use of inexpensive biomass byproducts, such as lignin, as carbon precursors in the production of hierarchical carbon materials for electrodes in electrochemical capacitors.

Insights into the Active Species of Nanoparticle-Functionalized Hierarchical Zeolites in Alkylation Reactions

Supported iron oxide nanoparticles have been incorporated onto hierarchical zeolites by microwave-assisted impregnation and mechanochemical grinding. Nanoparticle-functionalised porous zeolites were characterised by a number of analytical techniques such as XRD, N2 physisorption, TEM, and surface acidity measurements. The catalytic activities of the synthesised nanomaterials were investigated in an alkylation reaction. The results pointed to different species with varying acidity and accessibility in the materials, which provided essentially different catalytic activities in the alkylation of toluene with benzyl chloride under microwave irradiation, selected as the test reaction.

Capillary microreactors based on hierarchical SiO2 monoliths incorporating noble metal nanoparticles for the Preferential Oxidation of CO

Novel hierarchical SiO2 monolithic microreactors loaded with either Pd or Pt nanoparticles have been prepared in fused silica capillaries and tested in the Preferential Oxidation of CO (PrOx) reaction. Pd and Pt nanoparticles were prepared by the reduction by solvent method and the support used was a mesoporous SiO2 monolith prepared by a well-established sol–gel methodology. Comparison of the activity with an equivalent powder catalyst indicated that the microreactors show an enhanced catalytic behavior (both in terms of CO conversion and selectivity) due to the superior mass and heat transfer processes that take place inside the microchannel. TOF values at low CO conversions have been found to be ∼2.5 times higher in the microreactors than in the powder catalyst and the residence time seems to have a noticeable influence over the selectivity of the catalysts designed for this reaction. The Pd and Pt flexible microreactors developed in this work have proven to be effective for the CO oxidation reaction both in the presence and absence of H2, standing out as a very interesting and suitable option for the development of CO purification systems of small dimensions for portable and on-board applications.

Realizing the Commercial Potential of Hierarchical Zeolites: New Opportunities in Catalytic Cracking

Many approaches to mesoporous zeolites have been reported. The preparation of mesoporous zeolite Y, as the most widely used zeolite in catalysis, its properties, and its application in fluid catalytic cracking (FCC) and hydrocracking are reviewed. Finally, the scale-up and use of mesostrutured zeolite Y on an industrial scale are described, as the first commercial application of hierarchical zeolites.

Evaluation of sampling method effects in 3D non-rigid registration

Since the beginning of 3D computer vision problems, the use of techniques to reduce the data to make it treatable preserving the important aspects of the scene has been necessary. Currently, with the new low-cost RGB-D sensors, which provide a stream of color and 3D data of approximately 30 frames per second, this is getting more relevance. Many applications make use of these sensors and need a preprocessing to downsample the data in order to either reduce the processing time or improve the data (e.g., reducing noise or enhancing the important features). In this paper, we present a comparison of different downsampling techniques which are based on different principles. Concretely, five different downsampling methods are included: a bilinear-based method, a normal-based, a color-based, a combination of the normal and color-based samplings, and a growing neural gas (GNG)-based approach. For the comparison, two different models have been used acquired with the Blensor software. Moreover, to evaluate the effect of the downsampling in a real application, a 3D non-rigid registration is performed with the data sampled. From the experimentation we can conclude that depending on the purpose of the application some kernels of the sampling methods can improve drastically the results. Bilinear- and GNG-based methods provide homogeneous point clouds, but color-based and normal-based provide datasets with higher density of points in areas with specific features. In the non-rigid application, if a color-based sampled point cloud is used, it is possible to properly register two datasets for cases where intensity data are relevant in the model and outperform the results if only a homogeneous sampling is used.

The Benefits of Sampling Sports During Childhood

Sampling may promote prolonged engagement in sport by limiting physical injuries (Fraser-Thomas et al., 2005). Overtraining injuries are a concern for young athletes who specialize in one sport and engage in high volumes of deliberate practice (Hollander, Meyers, & Leunes, 1995; Law, Côté, & Ericsson, 2007). For instance, young gymnasts who practice for over 16 hours a week have been shown to have higher incidences of back injuries (Goldstein, Berger, Windier, & Jackson, 1991). A sampling approach in child-controlled play (e.g. deliberate play) rather than highly adult-controlled practice (e.g. deliberate practice) has been proposed as a strategy to limit overuse and other sport-related injuries (Micheli, Glassman, & Klein, 2000). In summary, sampling may protect against sport attrition by limiting sport related injuries and allowing children to have early experiences in sport that are enjoyable. Psychosocial Benefits of Sampling Only a small percentage of children who participate in school sports ever become elite athletes. Therefore, the psychosocial outcomes of sport participation are particularly important to consider. Recent studies with youth between the ages of 11 to 17 have found that those who are involved in a variety of extracurricular activities (e.g. sports, volunteer, arts) score more favourably on outcome measures such as Grade Point Average (GPA; Fredricks & Eccles, 2006a) and positive peer relationships (Fredricks & Eccles, 2006b) than youth who participate in fewer activities. These patterns are thought to exist due to each extracurricular activity bringing its own distinct pattern of socialization experiences that reinforce certain behaviours and/or teach various skills (Fredricks & Eccles, 2006b; Rose-Krasnor, Bussen, Willoughby, & Chambers, 2006). This contention is corroborated by studies of children and youths' experiences in extracurricular activities indicating that youth have unique experiences in each activity that contribute to their development (Hansen, Larson, & Dworkin, 2003; Larson, Hansen, & Moneta, 2006). This has led Wilkes and Côté (2007) to propose that children who sample different activities (through their own choice or by virtue of parental direction), have a greater chance of developing the following five developmental outcomes compared to children who specialize in one activity: 1) life skills, 2) prosocial behaviour, 3) healthy identity, 4) diverse peer groups and 5) social capital.

Core-shell nanocomposites prepared by hierarchical co-assembly: the role of the carbon shell in catalytic wet peroxide oxidation processes

The diffraction pattern of Fe3O4 (not shown) confirmed the presence of only one phase, corresponding to magnetite with a lattice parameter a = 8.357 Å and a crystallite size of 16.6 ± 0.2 nm. The diffraction pattern of MGNC (not shown) confirmed the presence of a graphitic phase, in addition to the metal phase, suggesting that Fe3O4 nanoparticles were successfully encapsulated within a graphitic structure during the synthesis of MGNC. The core-shell structure of MGNC is unequivocally demonstrated in the TEM micrograph shown in Fig. 1b. Characterization of the MGNC textural and surface chemical properties revealed: (i) stability up to 400 oC under oxidizing atmosphere; (ii) 27.3 wt.% of ashes (corresponding to the mass fraction of Fe3O4); (iii) a micro-mesoporous structure with a fairly well developed specific surface area (SBET = 330 m2 g-1); and (iv) neutral character (pHPZC = 7.1). In addition, the magnetic nature of MGNC (Fig. 2) is an additional advantage for possible implementation of in situ magnetic separation systems for catalyst recovery.

WGTTEST: Stata module to test the impact of sampling weights in regression analysis

wgttest performs a test proposed by DuMouchel and Duncan (1983) to evaluate whether the weighted and unweighted estimates of a regression model are significantly different.

Planktonic foraminifera of sediment trap samples

Biweekly sediment trap samples and concurrent hydrographic measurements collected between March 2005 and October 2008 from the Cariaco Basin, Venezuela, are used to assess the relationship between [CO3]2- and the area densities (ho A) of two species of planktonic foraminifera (Globigerinoides ruber (pink) and Globigerinoides sacculifer). Calcification temperatures were calculated for each sample using species-appropriate oxygen isotope (d18O) temperature equations that were then compared to monthly temperature profiles taken at the study site in order to determine calcification depth. Ambient [CO3]2- was determined for these calcification depths using alkalinity, pH, temperature, salinity, and nutrient concentration measurements taken during monthly hydrographic cruises. The rho A, which is representative of calcification efficiency, is determined by dividing individual foraminiferal shell weights (±0.43 µg) by their associated silhouette areas and taking the sample average. The results of this study show a strong correlation between rho A and ambient [CO3]2- for both G. ruber and G. sacculifer (R**2 = 0.89 and 0.86, respectively), confirming that [CO3]2- has a pronounced effect on the calcification of these species. Though the rho A for both species reveal a highly significant (p < 0.001) relationship with ambient [CO3]2-, linear regression reveals that the extent to which [CO3]2- influences foraminiferal calcification is species specific. Hierarchical regression analyses indicate that other environmental parameters (temperature and [PO4]3-) do not confound the use of G. ruber and G. sacculifer rho A as a predictor for [CO3]2-. This study suggests that G. ruber and G. sacculifer rho A can be used as reliable proxies for past surface ocean [CO3]2?-

Sediment and organic matter properties, and molecular formula from sampling locations in the Northern Black Sea

Remineralization of organic matter in reactive marine sediments releases nutrients and dissolved organic matter (DOM) into the ocean. Here we focused on the molecular-level characterization of DOM by high-resolution Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR-MS) in sediment pore waters and bottom waters from contrasting redox regimes in the northern Black Sea with particular emphasis on nitrogen-bearing compounds to derive an improved understanding of the molecular transformations involved in nitrogen release. The number of nitrogen-bearing molecules is generally higher in pore waters than in bottom waters. This suggests intensified degradation of nitrogen-bearing precursor molecules such as proteins in anoxic sediments: No significant difference was observed between sediments deposited under oxic vs anoxic conditions (average O/C ratios of 0.55) suggesting that the different organic matter quality induced by contrasting redox conditions does not impact protein diagenesis in the subseafloor. Compounds in the pore waters were on average larger, less oxygenated, and had a higher number of unsaturations. Applying a mathematical model, we could show that the assemblages of nitrogen-bearing molecular formulas are potential products of proteinaceous material that was transformed by the following reactions: (a) hydrolysis and deamination, both reducing the molecular size and nitrogen content of the products and intermediates; (b) oxidation and hydration of the intermediates; and (c) methylation and dehydration.