Pattern classification with mixtures of weighted least-squares support vector machine experts

Support Vector Machine (SVM) classifiers are high-performance classification models devised to comply with the structural risk minimization principle and to properly exploit the kernel artifice of nonlinearly mapping input data into high-dimensional feature spaces toward the automatic construction of better discriminating linear decision boundaries. Among several SVM variants, Least-Squares SVMs (LS-SVMs) have gained increased attention recently due mainly to their computationally attractive properties coming as the direct result of applying a modified formulation that makes use of a sum-squared-error cost function jointly with equality, instead of inequality, constraints. In this work, we present a flexible hybrid approach aimed at augmenting the proficiency of LS-SVM classifiers with regard to accuracy/generalization as well as to hyperparameter calibration issues. Such approach, named as Mixtures of Weighted Least-Squares Support Vector Machine Experts, centers around the fusion of the weighted variant of LS-SVMs with Mixtures of Experts models. After the formal characterization of the novel learning framework, simulation results obtained with respect to both binary and multiclass pattern classification problems are reported, ratifying the suitability of the novel hybrid approach in improving the performance issues considered.     

Robot task plan representation by Petri nets: modelling, identification, analysis and execution

In this paper we introduce a framework to represent robot task plans based on Petri nets. Our approach enables modelling a robot task, analysing its qualitative and quantitative properties and using the Petri net representation for actual plan execution. The overall model is obtained from the composition of simple models, leading to a modular approach. Analysis is applied to a closed loop between the robot controller and the environment Petri net models. We focus here on the quantitative properties, captured by stochastic Petri net models. Furthermore, we introduce a method to identify the environment and action layer parameters of the stochastic Petri net models from real data, improving the significance of the model. The framework building blocks and a single-robot task model are detailed. Results of a case study with simulated soccer robots show the ability of the framework to provide a systematic modelling tool, and of determining, through well-known analysis methods for stochastic Petri nets, relevant properties of the task plan applied to a particular environment.     

Electroluminescence of a device based on europium β-diketonate with phosphine oxide complex

Rare earth (RE) ions have spectroscopic characteristics to emit light in narrow lines, which makes RE complexes with organic ligands candidates for full color OLED (Organic Light Emitting Diode) applications. In particular, β-diketone rare earth (RE³⁺) complexes show high fluorescence emission efficiency due to the high absorption coefficient of the β-diketone and energy transfer to the central ion. In this work, the fabrication and the electroluminescent properties of devices containing a double and triple-layer OLED using a new β-diketone complex, [Eu(bmdm)₃(tppo)₂], as transporting and emitting layers are compared and discussed. The double and triple-layer devices based on this complex present the following configurations respectively: device 1: ITO/TPD (40 nm)/[Eu(bmdm)₃(tppo)₂] (40 nm)/Al (150 nm); device 2: ITO/TPD (40 nm)/[Eu(bmdm)₃(tppo)₂] (40 nm)/Alq₃ (20 nm)/Al (150 nm) and device 3: ITO/TPD (40 nm)/bmdm-ligand (40 nm)/Al (150 nm), were TPD is (N,N′-diphenyl-N,N′-bis(3-methylphenyl)-1,1-biphenil-4,4-diamine) and bmdm is butyl methoxy-dibenzoyl-methane. All the films were deposited by thermal evaporation carried out in a high vacuum system. These devices exhibit high intensity photo- (PL) and electro-luminescent (EL) emission. Electroluminescence spectra show emission from Eu³⁺ ions attributed to the ⁵D₀ to ⁷F_J (J = 0, 1, 2, 3 and 4) transitions with the hypersensitive ⁵D₀ → ⁷F₂ transition (around 612 nm) as the most prominent one. Moreover, a transition from ⁵D₁ to ⁷F₁ is also observed around 538 nm. The OLED light emission was almost linear with the current density. The EL CIE chromaticity coordinates (X = 0.66 and Y = 0.33) show the dominant wavelength, λ_d = 609 nm, and the color gamut achieved by this device is 0.99 in the CIE color space.     

Establishing quality performance of distribution companies based on yardstick regulation

This paper proposes a procedure and methodology for performance target setting related to continuity metrics in electricity distribution networks. This proposal deals with one of the key issues in the monopoly regulation, which is either the reduction of the information asymmetry among economic agents or the emulation of a competitive environment in industry segments usually considered as natural monopolies. This paper develops an approach toward standard definitions for Customer Average Interruption Duration Index (DEC) and Customer Average Interruption Frequency Index (FEC) based on inter-companies comparative analysis. This approach introduces productive efficiency concepts to the definition of network performance. In order to do so, the data envelopment analysis (DEA) technique, used for defining the efficiency frontier, was combined with the dynamic cluster technique, oriented toward models of identification of similar networks. A technique for comparative analysis was used for the definition of the network expected performance. Examples with the Brazilian electric systems are provided to clarify the methodology suggested by this paper.     

Spin Polarization by Tilted Magnetic Field in Wide Ga1?xAlxAs Parabolic Quantum Wells

Wide parabolic wells can be created by properly controlling the Al content during the growth of successive Ga_1–xAl_xAs thin layers. Under a tilted magnetic field these systems present interesting transport properties, which are associated to their composition dependent g-factor. We present an exact solution for the eigenstates of an electron gas inside such a quantum well. We calculate the renormalized cyclotron frequencies as functions of the angle of tilt, as well as the density of states, and the Fermi level. We discuss the conditions for the existence of spin-polarized charge.     

AC motor drive systems with a reduced-switch-count converter

This paper presents two reduced-switch-count AC drive systems. One of the configurations was conceived to operate with a two-phase motor and the other one with a three-phase motor. The drive systems provide both bidirectional power flow and power factor control. Pulsewidth modulation techniques for the converter control are discussed. Experimental results are presented.     

In vivo remineralization of acid-etched enamel in non-brushing areas as influenced by fluoridated orthodontic adhesive and toothpaste

This study aimed to evaluate the in vivo remineralization of acid-etched enamel in non-brushing areas as influenced by fluoridated orthodontic adhesive and toothpaste. One hundred and twenty teeth from 30 volunteers were selected. The teeth were assigned to four treatments: no treatment (negative control); 37% phosphoric acid-etching (PAE) (positive control); PAE + resin-modified glass ionomer cement (RMGIC); and, PAE + composite resin. Patients brushed teeth with fluoridated (n = 15) or non-fluoridated (n = 15) toothpastes, so that etched enamel was protected with screens and it was not in contact with the brush bristles. Remineralization was evaluated by means of laser fluorescence (LF), environmental scanning electronic microscopy, and energy dispersive spectrometry after extraction. The LF means were compared by means of Wilcoxon and Mann Whitney tests. Environmental scanning electron microscopy scores were compared among the groups using a Kruskal Wallis test, whereas the Ca/P ratio was evaluated by means of an Analysis of Variance with subparcels (treatments) and Tukey's post-hoc test. There were no statistically significant differences between the tooth pastes and between the orthodontic adhesives evaluated. Most teeth presented only partial enamel remineralization. Therefore, the fluoride released by the RMGIC was not enough to cause increased crystal regrowth in the acid-etched enamel. The use of fluoridated toothpaste did not provide positive additional effect.     

Stearoyl CoA Desaturase-1 Silencing in Glioblastoma Cells: Phospholipid Remodeling and Cytotoxicity Enhanced upon Autophagy Inhibition

Modulation of lipid metabolism is a well-established cancer hallmark, and SCD1 has been recognized as a key enzyme in promoting cancer cell growth, including in glioblastoma (GBM), the deadliest brain tumor and a paradigm of cancer resistance. The central goal of this work was to identify, by MS, the phospholipidome alterations resulting from the silencing of SCD1 in human GBM cells, in order to implement an innovative therapy to fight GBM cell resistance. With this purpose, RNAi technology was employed, and low serum-containing medium was used to mimic nutrient deficiency conditions, at which SCD1 is overexpressed. Besides the expected increase in the saturated to unsaturated fatty acid ratio in SCD1 silenced-GBM cells, a striking increase in polyunsaturated chains, particularly in phosphatidylethanolamine and cardiolipin species, was noticed and tentatively correlated with an increase in autophagy (evidenced by the increase in LC3BII/I ratio). The contribution of autophagy to mitigate the impact of SCD1 silencing on GBM cell viability and growth, whose modest inhibition could be correlated with the maintenance of energetically associated mitochondria, was evidenced by using autophagy inhibitors. In conclusion, SCD1 silencing could constitute an important tool to halt GBM resistance to the available treatments, especially when coupled with a mitochondria disrupter chemotherapeutic.     

Identification of Potential New Aedes aegypti Juvenile Hormone Inhibitors from N-Acyl Piperidine Derivatives: A Bioinformatics Approach

Aedes aegypti mosquitoes transmit several human pathogens that cause millions of deaths worldwide, mainly in Latin America. The indiscriminate use of insecticides has resulted in the development of species resistance to some such compounds. Piperidine, a natural alkaloid isolated from Piper nigrum, has been used as a hit compound due to its larvicidal activity against Aedes aegypti. In the present study, piperidine derivatives were studied through in silico methods: pharmacophoric evaluation (PharmaGist), pharmacophoric virtual screening (Pharmit), ADME/Tox prediction (Preadmet/Derek 10.0^®), docking calculations (AutoDock 4.2) and molecular dynamics (MD) simulation on GROMACS-5.1.4. MP-416 and MP-073 molecules exhibiting ΔG binding (MMPBSA −265.95 ± 1.32 kJ/mol and −124.412 ± 1.08 kJ/mol, respectively) and comparable to holo (ΔG binding = −216.21 ± 0.97) and pyriproxyfen (a well-known larvicidal, ΔG binding= −435.95 ± 2.06 kJ/mol). Considering future in vivo assays, we elaborated the theoretical synthetic route and made predictions of the synthetic accessibility (SA) (SwissADME), lipophilicity and water solubility (SwissADME) of the promising compounds identified in the present study. Our in silico results show that MP-416 and MP-073 molecules could be potent insecticides against the Aedes aegypti mosquitoes.