Generalized iterative RELIEF for supervised distance metric learning

The RELIEF algorithm is a popular approach for feature weighting. Many extensions of the RELIEF algorithm are developed, and I-RELIEF is one of the famous extensions. In this paper, I-RELIEF is generalized for supervised distance metric learning to yield a Mahananobis distance function. The proposed approach is justified by showing that the objective function of the generalized I-RELIEF is closely related to the expected leave-one-out nearest-neighbor classification rate. In addition, the relationships among the generalized I-RELIEF, the neighbourhood components analysis, and graph embedding are also pointed out. Experimental results on various data sets all demonstrate the superiority of the proposed approach.     

K-S检验与mRMR相结合的基因选择算法

为了解决基因数据集的基因选择难题,提出一种基于K-S检验与最小冗余最大相关（Minimum Redundancy-Maximum Relevance,mRMR）原则的基因选择算法。该算法先采用K-S检验选择出具有一定区分能力的基因,然后对选择到的基因进行mRMR判断,保留与类别高度相关而其间相关性较小的基因构成最终被选基因子集。以SVM为分类器,以F1_measure、分类准确率和AUC为评价指标对本文算法选择的基因子集进行评估,并将本文算法与K-S检验、mRMR,以及经典的RELIEF和FAST算法进行比较。5个经典基因数据集上的平均实验结果揭示：本文算法的运行时间远低于mRMR算法,且其各项评价指标值优于其他比较算法。因此,本文提出的K-S检验与mRMR结合的基因选择算法能选择到非常有效的基因子集。     

Flow transitions in vacuum arc remelting

Abstract

The metallurgical structure of an ingot produced by vacuum arc remelting (VAR) depends critically on the temperature distribution within the liquid portion of the partially solidified ingot. This, in turn, depends on the fluid motion in the pool, since the dominant mechanism for transporting heat is convection. There are three primary sources of motion: buoyancy; Lorentz forces arising from the passage of current through the pool; and Lorentz forces arising from the presence of external inductors. These forces are constantly in competition with each other, and each tends to induce a quite different distribution of velocity and temperature. We examine the transition between these different flow regimes and derive dimensionless criteria which determine which regime is dominant. We show that the structure of an ingot produced by VAR depends critically on the temperature distribution within the liquid portion of the partially solidified ingot. This, in turn, depends on the fluid motion in the pool, since the dominant mechanism for transporting heat is convection. There are three primary sources of motion: buoyancy; Lorentz forces arising from the passage of current through the pool; and Lorentz forces arising from the presence of external inductors. These forces are constantly in competition with each other, and each tends to induce a quite different distribution of velocity and temperature. We examine the transition between these different flow regimes and derive dimensionless criteria which determine which regime is dominant. We show that modest changes in ingot current can produce radical changes in temperature distribution, and that weak, steady magnetic fields, of only ～1 Gs, can induce a powerful swirling motion which suppresses the normal flow.     

Effects of hot deformation and accelerated cooling on microstructural evolution of low carbon microalloyed steels

Abstract

Using a Gleeble 1500 hot simulator, the effects of hot deformation parameters and accelerated cooling conditions on the microstructural characteristics of low carbon microalloyed steels were investigated by means of compression tests. It was found that the grain refinement effect of single pass reduction in the recrystallisation or unrecrystallisation temperature ranges is weaker than that of two pass reduction in the recrystallisation and unrecrystallisation temperature ranges. However, four pass deformation in the recrystallisation and unrecrystallisation temperature ranges could result in rather fine grained microstructures and, when coupled with moderately high cooling rate, partially acicular ferrite microstructure could be obtained. With the increase of cooling rate, the microstructure becomes finer and the content of acicular ferrite increases. Under similar deformation and cooling conditions, the specimens with relatively high carbon content have more refined microstructures.     

Process and compressive properties of porous nickel materials

Abstract

Compressive properties are investigated for the porous Ni materials processing by innovated powder metallurgical (PM) method. The porous Ni materials first show a short elastic region, then a long and oblique stress yield region within the strain range of about 10–50%, and finally, a densification region where the stress increases rapidly.     

Evaluation of performance of various color‐difference formulae using an experimental black dataset

The objective of this study was to develop a specific visual dataset comprising black‐appearing samples with low lightness (L* ranging from approximately 10.4 to 19.5), varying in hue and chroma, evaluating their visual differences against a reference sample, and testing the performance of major color difference formulas currently in use as well as OSA‐UCS‐based models and more recent CAM02 color difference formulas including CAM02‐SCD and CAM02‐UCS models. The dataset comprised 50 dyed black fabric samples of similar structure, and a standard (L*= 15.33, a* = 0.14, b* = ?0.82), with a distribution of small color differences, in ΔE^*_ab, from 0 to approximately 5. The visual color difference between each sample and the standard was assessed by 19 observers in three separate sittings with an interval of at least 24 hours between trials using an AATCC standard gray scale for color change, and a total of 2850 assessments were obtained. A third‐degree polynomial equation was used to convert gray scale ratings to visual differences. The Standard Residual Sum of Squares index (STRESS) and Pearson's correlation coefficient (r), were used to evaluate the performance of various color difference formulae based on visual results. According to the analysis of STRESS index and correlation coefficient results CAM02 color difference equations exhibited the best agreement against visual data with statistically significant improvement over other models tested. The CIEDE2000 (1:1:1) equation also showed good performance in this region of the color space. © 2013 Wiley Periodicals, Inc. Col Res Appl, 39, 589–598, 2014     

Effect of Structural Parameters on the Thermal Stress of a NiFe2O4-Based Cermet Inert Anode in Aluminum Electrolysis

Inert anode has been a hot issue in the aluminum industry for many decades. With the help of FEA(finite element analysis) software ANSYS,a model was developed to simulate the thermal stress distribution working condition of an inert anode. To reduce its thermal stress,the effect of some parameters on the thermal stress distribution was investigated,including the anode height,the anode radius,the hole depth,the hole radius,and the radius of inner chamfer and outer chamfer. The results showed that in the actual working condition of an inert anode,there existed a large axial tensile stress near the tangent interface between the anode and bath,which was the major cause of anode breaking. Increasing the anode height and reducing the hole depth properly seemed to be beneficial for the stress distribution. With the increase of anode radius,the stress distribution became better first and then deteriorated,the reasonable value was between 0.045 to 0.06m. The hole radius had a significant effect on the stress and a smaller radius would reduce the thermal stress. The effect of the radius of the inner chamfer and the outer chamfer was less than other parameters.     

3-D Finite Element Analysis of the Effect of Welding Residual Stress on Hydrogen Diffusion in Hydrogen Contained Environment

The hydrogen distribution of 16MnR steel weldment in hydrogen contained environment was calculated using the finite element method (FEM). The effect of welding residual stress on hydrogen diffusion has been discussed using a 3-D sequential coupling finite element analysis procedure complied by Abaqus code. The hydrogen diffusion coefficient in weld metal, the heat affected zone (HAZ), and the base metal of the 16MnR steel weldment were measured using the electrochemical permeation technique. The hydrogen diffusion without the effect of stress was also calculated and compared. Owing to the existence of welding residual stress, the hydrogen concentration was obviously increased and the hydrogen would diffuse and accumulate in the higher stress region.     

Optimal configuration of blast furnace slag runner to reduce fluid flow stresses at wall using mathematical modelling

Abstract

A three-dimensional mathematical model was developed to predict the wall shear stresses due to flow of liquid slag in slag runner of 'G' blast furnace of Tata Steel under different conditions. The liquid slag flow in the slag runner was considered to be turbulent and incompressible. The model was developed for single phase, steady state and isothermal conditions. To this end, the Navier Stokes equations along with continuity and turbulence equations (standard k–? model) were simultaneously solved with appropriate boundary conditions at the associated physical boundaries of the calculation domain. Several configurations were numerically assessed with respect to reduced shear stresses on the wall of the slag runner to select the best one. Due to accelerating flow the operating heights of liquid slag (density 2800 kg m^–3 at 1500°C) within the slag runner for different configurations were estimated with the help of Bernoulli's and continuity equations and fixed before the computation. The different configurations comprised of three segments with different parameters of either elevation or radius of curvature. Relatively high shear stresses were numerically predicted at the joint area of second and third segments of the slag runner for all the configurations. The radius of curvature was found as the dominant factor to reduce the shear stress at the joint region.     

Measuring and modelling residual stresses in butt welded P91 steel pipe including effects of phase transformations

Abstract

Residual stresses in a circumferentially butt welded steel pipe have been measured and numerically predicted. The pipe, containing the circumferential weld, has an outer diameter of 290 mm and a wall thickness of 55 mm, typical of components in power generation plants. An axisymmetric thermomechanical finite element (FE) simulation has been performed to obtain the residual stress field induced by the fusion welding of the pipe, taking solid state phase transformation effects into account and using temperature dependent material property data. Residual stresses have been measured using the X-ray diffraction and deep hole drilling techniques. Good correlation has been demonstrated with the predictions of the FE model. The paper demonstrates that a mixed experimental and numerical approach is useful for determining the residual stress distribution in welded joints.