We present shape measurements of Milky Way-sized dark matter halos at redshift z = 0 in a suite of 30 zoom simulations from the Auriga project. We compare the results in full magnetohydrodynamics against dark matter only simulations and find a strong influence of baryons in making dark matter haloes rounder at all radii compared to their dark matter only counterparts. At distances 30 kpc, rounder dark matter distributions correlate with extended massive stellar discs and low core gas densities. We measure the alignment between the halo and the disc shapes at different radii and find a high degree of alignment at all radii for most of the galaxies. In some cases the alignment significantly changes as a function of radius implying that the halo shape twists; this effect correlates with recently formed bulges and is almost absent in the dark matter only simulations. In a comparison against observational constraints we find that 20% of halos in our sample are consistent with observational results derived from the Pal 5 stream that favours an almost spherical shape. Including baryons is a required element to achieve this level of agreement. In contrast, none of the simulations (neither dark matter only nor with baryons) match the constraints derived from the Sagittarius stream that favour an oblate dark matter halo.

General noise cost functions have been recently proposed for support vector regression (SVR). When applied to tasks whose underlying noise distribution is similar to the one assumed for the cost function, these models should perform better than classical -SVR. On the other hand, uncertainty estimates for SVR have received a somewhat limited attention in the literature until now and still have unaddressed problems. Keeping this in mind, three main goals are addressed here. First, we propose a framework that uses a combination of general noise SVR models with naive online R minimization algorithm (NORMA) as optimization method, and then gives nonconstant error intervals dependent upon input data aided by the use of clustering techniques. We give theoretical details required to implement this framework for Laplace, Gaussian, Beta, Weibull and Marshall-Olkin generalized exponential distributions. Second, we test the proposed framework in two real-world regression problems using data of two public competitions about solar energy. Results show the validity of our models and an improvement over classical -SVR. Finally, in accordance with the principle of reproducible research, we make sure that data and model implementations used for the experiments are easily and publicly accessible.

No abstract

This paper describes the implementation of a methodology for classification of artificial lift system failures and adding a commonly used root cause failure classification from C-Fer (RIFT). The subject field is the oldest Colombian oilfield, which was discovered in 1918, and was produced by primary recovery, then by gas injection as secondary recovery; in the 80's the water flooding project started and finally, in 2006 optimization of water flooding took place increasing production from 4,500 bopd up to 40,000 bopd. Conditions are challenging for all types of artificial lift systems, including sand production, CO2 corrosion, H2S presence, high water cuts and unstable injection conditions. This methodology was applied to different artificial lift systems, such as: Beam Pump (BP), Progressive Cavity Pumps (PCP), Electrical Submersible Pumping Systems (ESP) and Electrical Submersible Progressive Cavity Pumping Systems (ESPCP). The starting point was the definition or limitation of the boundaries of every system, since some of them use rods for power transmission and others use a power cable. Then, every job in the well was defined as Failure, Failure No-ALS or No Failure in order to classify and look for ways to improve run life and decrease failures by implementing action plans with operations, engineering, and suppliers among others. The addition of root cause failure classification helped to refine the action plans in order to solve the issues. In some cases, it was necessary to customize RIFT classification in order to track local issues; such as: unstable injection wells, drilling operations in the same well location, etc (some examples are included). Included in this methodology, some KPIs: failure index, indirect failure index, pulling index, recurrence index, average run life, average run time were used. These KPIs helped track performance improvement in the last two years, getting excellent results.

This paper presents the graffiti interpretation of one-stroke handwritten digits (0 to 9) and commands (backspace, carriage return and space) using support vector machines (SVMs). A number of SVM-based graffiti interpreters are proposed for the recognition of graffiti. The performance of the proposed SVM-based graffiti interpreters subject to various kernel functions and parameters are investigated. Simulation and experimental results are presented to show the applicability and the merits of various graffiti interpreters.