Search

There are currently many barriers that prevent non-experts from exploiting machine learning solutions ranging from the lack of intuition on statistical learning techniques to the trickiness of hyperparameter tuning. Such barriers have led to an explosion of interest in automated machine learning (AutoML), whereby an off-the-shelf system can take care of many of the steps for end-users without the need for expertise in machine learning. This paper presents Ensemble Squared (Ensemble$^2$), an AutoML system that ensembles the results of state-of-the-art open-source AutoML systems. Ensemble$^2$ exploits the diversity of existing AutoML systems by leveraging the differences in their model search space and heuristics. Empirically, we show that diversity of each AutoML system is sufficient to justify ensembling at the AutoML system level. In demonstrating this, we also establish new state-of-the-art AutoML results on the OpenML tabular classification benchmark. more...

Background: Human papilloma virus (HPV) infection is one of the most common sexually transmitted diseases that affects men like women and infected cutaneous and mucosal squamous epithelium. The aim of the present study was to determine the prevalence of HPV in the semen of oligospermic, azoospermic and normal patients. Methods: From June 2012 to June 2013, a total of 90 individuals were enrolled in this cross sectional comparative study. The participants were classified into three groups (oligospermia, azoosprmia and normal). This classification was based on a new WHO reference values for human semen characteristics published on 2010. After extraction of DNA from specimens L1 gene of HPV was amplified by nested polymerase chain reaction (Nested-PCR) and the PCR products of positive specimens were genotyped using INNO-LiPA HPV Genotyping Extra assay. Results: Among 50 confirmed oligospermic male, 15 were HPV DNA positive (30%). In azoospemic group we had 8 HPV DNA positive (40%) and in normal group just 3 of 20(15%) samples were positive. Statistical assessment was done with SPSS v.15. Chi-square test showed no significant relationship between 3 groups results. Based on independent samples t-test, we found statistical significant relationship for sperm count (p more...

"nBackground: Methyl tert-butyl ether (MTBE) has been commercially used as an octane enhancer to replace tetraethyl lead in gasoline since 1979. The high mobility, water solubility, and resistance to natural attenuation associated with MTBE may re­sult in contamination of ground and surface waters. In this investigation the degradation of aqueous MTBE at relatively high concentrations was studied by UV-vis/TiO2/O2 photocatalytic process. The effect of important operational parameters such as pH, oxygen flow, catalyst loading, and irradiation time were also studied."nMethods:Concentration of MTBE and intermediates such as tert-butyl formate (TBF) and tert-butyl alcohol (TBA) were meas­ured using a gas chromatograph equipped with flam ionization detector and combined with headspace sampler."nResults: The time required for complete degradation increased from 15 to 150 min, when the initial concentration was in­creased from 10 to 500 mg/L. The first order rate constant for degradation of MTBE from the hydroxyl radical was esti­mated to be 0.266 to 0.033 min-1 as the concentration increased from 10 to 500 mg/L. Study on the overall mineralization moni­tored by total organic carbon (TOC) analysis showed that in the initial concentration of 100 mg/L MTBE, complete min­eralization was obtained after 110 min under UV-vis/TiO2/O2 photocatalytic process."nConclusion: The data presented in this paper clearly indicate that UV/TiO2/O2 advanced oxidation process provides an effi­cient treatment alternative for the remediation of MTBE contaminated water. &nbsp more...

Landfill leachate is one of the most important sources of toxic organic compounds for ground and surface waters. Advanced oxidation processes can offer an effective and environmentally friendly method for pretreatment of landfill leachates. In this study, an ultrasonic process was used for the pre-treatment of landfill leachate with the objective of improving its overall biodegradability, evaluated in terms of the BOD5/COD ratio, up to a value compatible with biological treatment. Under optimized experimental conditions (pH of 10, power of 110 watts, frequency of 60 kHz, TiO2 concentration of 5 mg/L and exposure time of 120 min), this method showed suitability for partial removal of chemical oxygen demand (COD). The biodegradability was significantly improved (BOD5/COD increased from 0.210 to 0.786) which allowed an almost total removal of COD by a sequential activated sludge process. more...

Recent studies on heterostructures of ultrathin ferromagnets sandwiched between a heavy metal layer and an oxide have highlighted the importance of spin-orbit coupling (SOC) and broken inversion symmetry in domain wall (DW) motion. Specifically, chiral DWs are stabilized in these systems due to the Dzyaloshinskii-Moriya interaction (DMI). SOC can also lead to enhanced current induced DW motion, with the Spin Hall effect (SHE) suggested as the dominant mechanism for this observation. The efficiency of SHE driven DW motion depends on the internal magnetic structure of the DW, which could be controlled using externally applied longitudinal in-plane fields. In this work, micromagnetic simulations and collective coordinate models are used to study current-driven DW motion under longitudinal in-plane fields in perpendicularly magnetized samples with strong DMI. Several extended collective coordinate models are developed to reproduce the micromagnetic results. While these extended models show improvements over traditional models of this kind, there are still discrepancies between them and micromagnetic simulations which require further work. more...

The main analytical model for describing the motion of magnetic domain walls is the 1-D model formulated based on the profile of a Bloch wall. This model qualitatively describes the motion of magnetic domain wall in nanowires, while it may fail to match experimental and numerical results quantitatively. In recent years, the 1-D model has been further generalized by the introduction of terms such as spin transfer torques and spin orbit torques. It has also been used to describe the motion of different domain walls, including vortex walls. It seems that in many such attempts, formalisms are not followed accurately and the main assumptions of the model (such as the Bloch wall profile used in developing the model) are underestimated. In this paper, we first derive an analytical model to describe the motion of a tilting Bloch wall in perpendicularly magnetized materials using four collective coordinates. We then compare the energy landscape predicted by this model to that of micromagnetic simulations, highlighting the possibility of using such comparisons to develop corrections for the 1-D model. more...