Model Change Active Learning in Graph-Based Semi-supervised Learning

Active learning in semi-supervised classification involves introducing additional labels for unlabelled data to improve the accuracy of the underlying classifier.A challenge is to identify which points to label to best improve performance while limiting the number of new labels."Model Change"active learning quantifies the resulting change incurred in the classifier by introducing the additional label(s).We pair this idea with graph-based semi-supervised learning(SSL)methods,that use the spectrum of the graph Laplacian matrix,which can be truncated to avoid prohibitively large computational and storage costs.We consider a family of convex loss functions for which the acquisition function can be efficiently approximated using the Laplace approximation of the posterior distribution.We show a variety of multiclass examples that illustrate improved performance over prior state-of-art.

Risk Assessment of Diary Cattle Brucellosis Based on Systematic Multilevel Grey Relation Entropy Method

[Objective] The study was to explore the major factors affecting diary cattle brucellosis risk assessment,as well as their strong-to-weak sequence,so as to provide theoretical basis for assessing diary cattle brucellosis risk level in different regions.[Method] From 4 dimensions of feeding and importing,breeding,housing and polyculture situation,an evaluation index system was set up,and diary cattle brucellosis risk survey was conducted in 3 typical regions.Finally,systematic multilevel grey relation entropy method was applied to perform data analysis.[Result] The strong-to-weak sequence of Level 1 impact factor of diary cattle brucellosis was as follows：feeding and importinghousingpolyculture situationbreeding;the sequence of Level 2 impact factor was U32〉U12〉U11〉U31〉U21〉U42〉U43〉U23〉U22〉U41;the risk level sequence of 3 typical regions was Province A（County A1,A2,A3）Province B（County B1,B2,B3）Province C（County C1,C2,C3）.[Conclusion] According to the weight of Level 1 index strata,administrative departments at all levels and dairy cattle farmers should lay emphasis on the aspects of feeding,importing and housing;viewed from the perspective of Level 2 index strata,dairy cattle farmers should value the siting of cattle field,the brucellosis surveillance before importing and milking modes most.According to the diary cattle brucellosis risk level of 3 typical regions,if administrative departments at all levels strengthen peoples＇ awareness of their personal health and increase investment in this area,with new healthy cultured atmosphere built,the risk level of diary cattle brucellosis will surly decline.

Response surface method using grey relational analysis for decision making in weapon system selection

A proper weapon system is very important for a na- tional defense system. Generally, it means selecting the optimal weapon system among many alternatives, which is a multiple- attribute decision making （MADM） problem. This paper proposes a new mathematical model based on the response surface method （RSM） and the grey relational analysis （GRA）. RSM is used to obtain the experimental points and analyze the factors that have a significant impact on the selection results. GRA is used to an- alyze the trend relationship between alternatives and reference series. And then an RSM model is obtained, which can be used to calculate all alternatives and obtain ranking results. A real world application is introduced to illustrate the utilization of the model for the weapon selection problem. The results show that this model can be used to help decision-makers to make a quick comparison of alternatives and select a proper weapon system from multiple alternatives, which is an effective and adaptable method for solving the weapon system selection problem.

Program evaluation and its application to equipment based on super-efficiency DEA and gray relation projection method

For the gray attributes of the equipment program and its difficulty to carry out the quantitative assessment of the equipment program information, the gray relation projection method is simply reviewed. Combining the super-data envelopment analysis（DEA） model and the gray system theory, a new super-DEA for measuring the weight is proposed, and a gray relation projection model is established to rank the equipment programs. Finally, this approach is used to evaluate the equipment program. The results are verified valid and can provide a new way for evaluating the equipment program.

Optimization of cutting parameters with Taguchi and grey relational analysis methods in MQL-assisted face milling of AISI O2 steel

This study aims to examine the usability of environmentally harmless vegetable oil in the minimum quantity of lubrication(MQL)system in face milling of AISI O2 steel and to optimize the cutting parameters by different statistical methods.Vegetable oil was preferred as cutting fluid,and Taguchi method was used in the preparation of the test pattern.After testing with the prepared test pattern,cutting performance in all parameters has been improved according to dry conditions thanks to the MQL system.The highest tool life was obtained by using cutting parameters of 7.5 m cutting length,100 m/min cutting speed,100 mL/h MQL flow rate and 0.1 mm/tooth feed rate.Optimum cutting parameters were determined according to the Taguchi analysis,and the obtained parameters were confirmed with the verification tests.In addition,the optimum test parameter was determined by applying the gray relational analysis method.After using ANOVA analysis according to the measured surface roughness and cutting force values,the most effective cutting parameter was observed to be the feed rate.In addition,the models for surface roughness and cutting force values were obtained with precisions of 99.63%and 99.68%,respectively.Effective wear mechanisms were found to be abrasion and adhesion.

Convergence analysis for the Secant method based on new recurrence relations

A new convergence theorem for the Secant method in Banach spaces based on new recurrence relations is established for approximating a solution of a nonlinear operator equation. It is assumed that the divided difference of order one of the nonlinear operator is Lipschitz continuous. The convergence conditions differ from some existing ones and are easily satisfied. The results of the paper are justified by numerical examples that cannot be handled by earlier works.

An ideal point method adopting grey relational entropy to estimate life cycle cost with effectiveness of torpedo

To making the decision of the developing blue prints,ideal point method was selected to estimate the life cycle cost with effectiveness of torpedo.At the same time,the concept of grey relational entropy of the grey system theory was adopted to compute the distance between each blue print and the ideal point(or negative ideal point).The blue print,nearest to the ideal point and farthest to the negative ideal point,is the best one.As an example,four blue prints of torpedo were estimated.The result indicates the practical value of this method.

Priority-sequence of mineral resources' development and utilization based on grey relational analysis method

Generally, the sequence decision of the development and utilization of Chinese mineral resources is based on national and provincial overall plan of the mineral resources. Such plan usually cannot reflect the relative size of the suitability of the development and utilization of mineral resources. To solve the problem, the paper has selected the gift condition, the market condition, the technological condition,socio-economic condition and environmental condition as the starting-points to analyze the influential factors of the priority-sequence of mineral resources' development and utilization. The above 5 conditions are further specified into 9 evaluative indicators to establish an evaluation indicator system. At last,we propose a decision model of the priority sequence based on grey relational analysis method, and figure out the observation objects by the suitability index of development. Finally, the mineral resources of a certain province in China were analyzed as an example. The calculation results indicate that silver(2.0057), coal(1.9955), zinc(1.9442), cement limestone(1.9077), solvent limestone(1.5624) and other minerals in the province are suitable for development and utilization.

New Refinement Relations of Z Specifications for Multiple Viewpoints Oriented Requirements Method

In this paper we develop several new refinement relations of Z for multiple viewpoints oriented requirements method (MVORM). The original motivation is that we found the standard Z refinement relation is not adequate or correct when considering specifications that have temporal relationships of operations. The concept of temporal state variables is introduced into Z. Then new implementation relations are defined and new refinement relations are deduced, mainly for temporal state variables to process temporal relationships of operations. We use state transition systems to abstract the temporal state transitions. A simple example is used to show the procedures of MVORM. Finally some directions of further work are forwarded.

Optimization of Tribological Performance of Al-6061T6-15% SiCp-15% Al₂O₃Hybrid Metal Matrix Composites Using Taguchi Method &Grey Relational Analysis

In this investigation, optimization of tribological performance parameters of Al-6061T6 alloy reinforced with SiC (15% by weight) and Al2O3 (15% by weight) particulates having particle size of 37 μm each has been presented. The wear and frictional properties of the hybrid metal matrix composites have been studied by performing dry sliding wear test using pin-on-disc wear tester. A L27 orthogonal array is selected for the analysis of the data. From the test results it is observed that sliding distance has the significant contribution in controlling the friction and wear behaviour of hybrid composites. A confirmation test is also carried out to verify the accuracy of the results obtained through the optimization. In addition an optical micrograph test is also performed on the wear tracks to study the wear mechanism.

Optimizing the Glass Fiber Cutting Process Using the Taguchi Methods and Grey Relational Analysis

This paper mainly describes a new approach to optimizing of the cutting glass fiber with multiple performance characteristics, based on reliability analysis, Taguchi and Grey methods. During the cutting process, the speed, the volume and the cutting load are optimized cutting parameters when the performance characteristics, which include Weibull modulus and blade wear, are taken into consideration. In this paper, optimization with multiple performance characteristics is found to be the highest cutting speed and the smallest cutting volume, and the medium cutting load. An analysis of the variance of the blade wear indicates that the cutting speed (47.21%), the cutting volume (14.62%) and the cutting load (12.20%) are the most significant parameters in the cutting process of glass fibers. In summary, the most optimal cutting parameter should be A3B1C2. The results of experiments have shown that the multiple performance characteristics of cutting glass fiber are improved effectively through this approach.

ON RELATIONS BETWEEN THE MODIFIED-ITERATIVE METHOD AND CHIEN’S PERTURBATION SOLUTION

In this paper, we gave analytical formulas of characteristic relation of circular plate in solving high-order solutions of modified-iterative method, which reduces the calculating quantities of the method. Having deduced the relations between the modified-iterative method and Chien's perturbation solution, we obtained the conclusion that the convergent regions of the two methods are the same.

Analysis on decision-making model of plan evaluation based on grey relation projection and combination weight algorithm

In military service joint operations, when there are more operational forces, more multifarious materials are consumed, the support is more complex and fuzzy, the deployment of personnel is more rapid, and the support provided by wartime military material support powers can be more effective. When the principles,requirements, influencing factors and goals of military material support forces are deployed in wartime, an evaluation indicator system is established. Thus, a new combined empowerment method based on an analytic hierarchy process（AHP） is developed to calculate the subjective weights, and the rough entropy method is used to calculate the objective weights. Combination weights can be obtained by calculating the weight preference coefficient error, which is determined by combining the cooperative game method and the minimum deviation into objectives. This approach can determine the grey relation projection coefficient and synthesize the measure scheme superiority to finally optimize the deployment plan using the grey relation projection decision-making method. The results show that the method is feasible and effective;it can provide a more scientific and practical decision-making basis for the military material support power deployment in wartime.

Particle characteristics and rheological constitutive relations of high concentration red mud

Red mud has relatively small solid particles （d50= 13.02 μm） and will flow in paste form under high pressure during pipeline transport. Red mud belongs to a two-phase flow of materials with high viscosity and a high concentration of non-sedimentation, homogeneous solid-liquids. It is difficult to test its rheological properties under atmospheric pressure. Measurements such as rotational viscometry can not reflect the real state of the material when it is flowing in a pipe. Tested rheological parameters are somewhat higher than the actual values. In our investigation, grain shape, distinctive modality and grain size distribution of red mud were tested. Based on the principle of tube measurement, rheological experiments on red mud at different concentrations were carried out by using our independently developed tube-type pressure theology test facility, and obtained constitutive equations. We conclude that red mud behaves as non-Newtonian pseudo-plastic fluid in pipe flows. Its consistency and power-law indices vary considerably with different concentrations.

Poisson theory and integration method for a dynamical system of relative motion

This paper focuses on studying the Poisson theory and the integration method of dynamics of relative motion. Equations of a dynamical system of relative motion in phase space are given. Poisson theory of the system is established. The Jacobi last multiplier of the system is defined, and the relation between the Jacobi last multiplier and the first integrals of the system is studied. Our research shows that for a dynamical system of relative motion, whose configuration is determined by n generalized coordinates, the solution of the system can be found by using the Jacobi last multiplier if （2n-1） first integrals of the system are known. At the end of the paper, an example is given to illustrate the application of the results.

A pseudoenergy wave-activity relation for ageostrophic and non-hydrostatic moist atmosphere

By employing the energy-Casimir method, a three-dimensional virtual pseudoenergy wave-activity relation for a moist atmosphere is derived from a complete system of nonhydrostatic equations in Cartesian coordinates. Since this system of equations includes the effects of water substance, mass forcing, diabatic heating, and dissipations, the derived wave-activity relation generalizes the previous result for a dry atmosphere. The Casimir function used in the derivation is a monotonous function of virtual potential vorticity and virtual potential temperature. A virtual energy equation is employed （in place of the previous zonal momentum equation） in the derivation, and the basic state is stationary but can be three-dimensional or, at least, not necessarily zonally symmetric. The derived wave-activity relation is further used for the diagnosis of the evolution and propagation of meso-scale weather systems leading to heavy rainfall. Our diagnosis of two real eases of heavy precipitation shows that positive anomalies of the virtual pseudoenergy wave-activity density correspond well with the strong precipitation and are capable of indicating the movement of the precipitation region. This is largely due to the cyclonic vorticity perturbation and the vertically increasing virtual potential temperature over the precipitation region.

Modified theoretical stage-discharge relation for circular sharp-crested weirs

A circular sharp-crested weir is a circular control section used for measuring flow in open channels, reservoirs, and tanks. As flow measuring devices in open channels, these weirs are placed perpendicular to the sides and bottoms of straight-approach channels. Considering the complex patterns of flow passing over circular sharp-crested weirs, an equation having experimental correlation coefficients was used to extract a stage-discharge relation for weirs. Assuming the occurrence of critical flow over the weir crest, a theoretical stage-discharge relation was obtained in this study by solving two extracted non-linear equations. To study the precision of the theoretical stage-discharge relation, 58 experiments were performed on six circular weirs with different diameters and crest heights in a 30 cm-wide flume. The results show that, for each stage above the weirs, the theoretically calculated discharge is less than the measured discharge, and this difference increases with the stage. Finally, the theoretical stage-discharge relation was modified by exerting a correction coefficient which is a function of the ratio of the upstream flow depth to the weir crest height. The results show that the modified stage-discharge relation is in good agreement with the measured results.

Dynamic Stiffness Deterioration of a Machining Center Based on Relative Excitation Method

The tool point frequency response function(FRF) is commonly obtained by impacting test or semi-analytical techniques.Regardless of the approach,it is assumed that the workpiece system is rigid.The assumption is valid in common machining,but it doesn’t work well in the cutting processes of thin-wall products.In order to solve the problem,a multi-degree-of-freedom dynamic model is employed to obtain the relative dynamic stiffness between the cutting tool and the workpiece system.The relative direct and cross FRFs between the cutting tool and workpiece system are achieved by relative excitation experiment,and compared with the tool point FRFs at x and y axial direction.The comparison results indicate that the relative excitation method could be used to obtain the relative dynamic compliance of machine-tool-workpiece system more actually and precisely.Based on the more precise relative FRFs,four evaluation criterions of dynamic stiffness are proposed,and the variation trend curves of these criterions during the last six months are achieved and analyzed.The analysis results show that the lowest natural frequency,the maximum and the average dynamic compliances at x axial direction deteriorate more quickly than that at y axial direction.Therefore,the main cutting direction and the large-size direction of workpieces should be arranged at y axial direction to slow down the deterioration of the dynamic stiffness of machining centers.The compliance of workpiece system is considered,which can help master the deterioration rules of the dynamic stiffness of machining centers,and enhance the reliability of machine centers and the consistency of machining processes.

DYNAMICAL STABILITY OF VISCOELASTIC COLUMN WITH FRACTIONAL DERIVATIVE CONSTITUTIVE RELATION

The dynamic stability of simple supported viscoelastic column, subjected to a periodic axial force, is investigated. The viscoelastic material was assumed to obey the fractional derivative constitutive relation. The governing equation of motion was derived as a weakly singular Volterra integro-partial-differential equation, and it was simplified into weakly singular Volterra integro-ordinary-differential equation by the Galerkin method. In terms of the averaging method, the dynamical stability was analyzed. A new numerical method is proposed to avoid storing all history data. Numerical examples are presented and the numerical results agree with the analytical ones.

Adjoint Method-Based Algorithm for Calculating the Relative Dispersion Ratio in a Hydrodynamic System

Relative dispersion ratio(RDR)can be used to quantify the deviation behavior of a water parcel’s trajectory caused by a disturbance in a hydrodynamic system.It can be calculated by using a standard method for determining relative dispersion(RD),which accounts for the growth of the deviation of a cluster of particles from a specific initial time.However,the standard method for computing RD is time consuming.It involves numerous computations on tracing many water parcels.In this study,a new method based on the adjoint method is proposed to acquire a series of RDR fields in one round of tracing.Through this method,the continuous variation in the RDR corresponding to a time series of the disturbance time t can be obtained.The consistency and efficiency of the new method are compared with those of the standard method by applying it to a double-gyre flow and an unsteady Arnold-Beltrami-Childress flow field.Results show that the two methods have good consistency in a finite time span.The new method has a notable speedup for evaluating the RDR at multiple t.