商标图像检索中子图像特征融合准则研究

该文对商标图像检索中的子图像特征融合准则进行了研究 ,根据子图像可信度对子图像特征融合准则进行加权改进。用Hu不变矩进行基于子图像特征融合的商标图像检索实验 ,用PVR指数作为图像检索性能评价准则。实验表明 ,基于子图像特征融合的商标图像检索优于基于全局图像特征的检索 ,子图像特征融合加权准则优于未经加权的子图像特征融合准则 。

基于子空间信息量准则的软测量模型选择研究

提出了子空间信息量概念及其准则,并引出了函数集信息量的概念;在此基础上,阐述了基于上述概念和准则的软测量模型选择理论,指出并解决了软测量模型选择欠学习与过学习问题;给出了一种次优软测量模型选择算法。最后,通过具体实例验证了上述方法的可行性和优越性。

应用子空间信息准则选择模型参数

指出子空间信息准则是模型选择的一种新准则,它在一些假设条件下,给出推广误差的一种无偏估计.研究再生核Hilbert空间中,应用子空间信息准则于模型参数选择问题,证明子空间信息准则是推广误差的一种无偏估计,同时说明学习算法有较好的推广能力.

论如何提高子空间信息准则的精度

在机器学习中,子空间信息准则给出推广误差的一种无偏估计.本文分析降低SIC精度的原因,并且就如何提高SIC精度进行探讨.

基于子空间信息量准则的前向神经网络设计理论与方法研究

从理论上提出了子空间信息量(SIQ)及其准则(SIQC)的概念;在此基础上阐述了基于上述准则的前向神经网络设计的相关理论,包括前向神经网络隐含层信息量(HL IQ)、存在性和逼近定理,给出了选择隐含层神经元数、权值向量集和隐含层激励函数的指导方向;提出了基于上述理论的一种可行的次优网络设计算法;最后,详细分析了网络性能指标及其影响因素.上述理论和方法完全克服了传统学习算法的各种弊端,丰富了前向神经网络设计领域的理论依据,具有较大的理论指导和实际应用价值.文中通过具体实例验证了上述理论和方法的可行性和优越性.

应用子空间信息准则选择模型参数

在监督学习中,模型的选择直接影响学习算法的推广能力.子空间信息准则是模型选择的一种新准则,它在一些假设条件下,给出推广误差的一种无偏估计.研究再生核Hilbert空间时,将子空间信息准则应用于模型参数选择问题,也证明子空间信息准则是推广误差的一种无偏估计.

网络管理之子网划分准则

网络中的问题是不计其数的，然而如果网络在一开始时就设计和安装恰当的话，那么大多数的问题都是可以避免的。

运用合作原则中准则的违背来加深听力理解

CP(cooperative principle)原则是会话含义理论中的一条重要原则,然而在真实的话语交际中,人们确实对其中的子准则(maxim)会有所违背,而说话者的真实意图,只有依赖合作原则才能推导出来。对于英语听力测试中的受试者,只有具备了相应的认知语境和知识结构,并运用这一原则,才能更有效地领悟会话的含义,从而加深自己的听力理解。

神经网络过学习问题的统计学分析及改进算法

着重分析了前向神经网络在学习过程中过学习问题产生的统计学机理 ,并根据统计学习理论提出了一种新的准则函数·该准则函数综合考虑了网络的逼近性、容错性和归纳性 ,并且算法实现起来简单 ,易于在线选择·同时还提出一种新的网络结构学习算法·最后用该学习算法进行网络训练 ,仿真结果显示了算法的有效性和实用性·

一种基于子空间距离准则的源数目估计算法

源数目估计是阵列信号处理中的一个重要课题,也是声纳探测领域的一个难题。通过研究特征子空间分析方法,提出了一种新的源数目估计算法。该方法在盲源分离结果的基础上,基于信号子空间和噪声子空间正交的原理,建立了子空间代价函数来估计观测空间的源数目。同时,通过仿真和水池实验,对该方法的性能做了研究,并与基于信息论准则中最小长度描述准则的源数目估计算法进行了比较。仿真和水池实验结果表明,该源数目估计算法能得到准确的源数目,并具有较好的抗噪性,在水声环境中具有更好的稳健性。

Stress-drop effect on brittleness evaluation of rock materials

Uniaxial or triaxial compression test of cylindrical rock specimens using rock mechanics testing machine is a basic experimental method to study the strength and deformation characteristics of rock and the development process of rock fracture. Extensive literature review has been conducted on this issue;meanwhile, experimental and numerical studies have been conducted on the stress-drop effect on the brittleness of rock materials. A plastic flow factor of λ is proposed to describe the stress-drop effect. Evaluation methods of the factor λ corresponding to the four yield criteria of rock mass are proposed. Those four yield criteria are Tresca criterion, von-Mises criterion, Mohr-Coulomb criterion and Drucker-Prager criterion. For simplicity purposes, an engineering approximation approach has been proposed to evaluate the stress-drop with a non-zero strain increment. Numerical simulation results validated the effectiveness of the plastic flow factors λ as well as the engineering approximation approach. Based on the results in this study, finite element code can be programmed for brittle materials with stress-drop, which has the potential to be readily incorporated in finite element codes.

Upper bound solution of surrounding rock pressure of shallow tunnel under nonlinear failure criterion

The nonlinear Baker failure criterion is introduced into the upper-bound limit analysis to examine the face stability of a shallow tunnel. The tunnel face under the ultimate limit state is analyzed from the perspective of energy balance. The work rates of external forces and internal energy dissipation are calculated. An analytical solution of necessary face pressures is derived. The optimal upper-bound solution of the face pressures is obtained by optimization. The results show that the three dimensionless parameters A, T, n of nonlinear Baker failure criterion have different effects on the necessary face pressures and the pattern failure mechanisms ahead of tunnel face. A is the most important one;n takes the second place, and T is the least one. The computed necessary face pressures are nonlinearly increasing when A is reduced. Combined with the actual monitoring data of Taxia tunnel, the calculation results in this paper is verified. It is suggested that the tunnel face supports should be strengthened timely in soft rocks to prevent the occurrence of face collapse.

Thermo-hydro-mechanical-chemical(THMC)coupling fracture criterion of brittle rock

Based on analysis of thermo-hydro-mechanical-chemical(THMC)coupling mechanism for brittle rock,THMC coupling indicator in terms of rock porosity was introduced to represent the influencing degree of THMC coupling field on stress field in order to establish THMC coupling fracture criterion.A novel real-time measurement method of permeability(related to porosity)was proposed to determine the THMC coupling indicator,and self-designed THMC coupling tests and scanning electron microscope tests were conducted on pre-cracked red sandstone specimens to study the macroscopic and microscopic fracture mechanism.Research results show that the higher the hydraulic pressure is,the smaller the crack initiation load is and the easier the Mode I fracture occurs.Test results are in good agreement with prediction results(crack initiation load and angle,and fracture mode),which can verify the effectiveness of the newly established THMC coupling fracture criterion.This new fracture criterion can be also further extended to predict THMC coupling fracture of multi-crack problem.

An ICPSO-RBFNN nonlinear inversion for electrical resistivity imaging

To improve the global search ability and imaging quality of electrical resistivity imaging(ERI) inversion, a two-stage learning ICPSO algorithm of radial basis function neural network(RBFNN) based on information criterion(IC) and particle swarm optimization(PSO) is presented. In the proposed method, IC is applied to obtain the hidden layer structure by calculating the optimal IC value automatically and PSO algorithm is used to optimize the centers and widths of the radial basis functions in the hidden layer. Meanwhile, impacts of different information criteria to the inversion results are compared, and an implementation of the proposed ICPSO algorithm is given. The optimized neural network has one hidden layer with 261 nodes selected by AKAIKE's information criterion(AIC) and it is trained on 32 data sets and tested on another 8 synthetic data sets. Two complex synthetic examples are used to verify the feasibility and effectiveness of the proposed method with two learning stages. The results show that the proposed method has better performance and higher imaging quality than three-layer and four-layer back propagation neural networks(BPNNs) and traditional least square(LS) inversion.

Entanglement Criterion of N-Qubit State

In this paper, we shall develop a generic scheme to construct the criterion that an N-qubit state has true N-particle entanglement. For the N = 3 case, we show that violation of the Mermin＇s inequality │ E^LHV （σ^ x× σ^ y× σ^y ＋ σ^y × σ^x×σ^y ＋ σ^y ×σ^y× σ^x＆ -σ^x × σ^x ×σ^x）│≤2, is sufficient to confirm three-particle entanglement.

A K-means clustering based blind multiband spectrum sensing algorithm for cognitive radio

In this paper,a blind multiband spectrum sensing(BMSS)method requiring no knowledge of noise power,primary signal and wireless channel is proposed based on the K-means clustering(KMC).In this approach,the KMC algorithm is used to identify the occupied subband set(OSS)and the idle subband set(ISS),and then the location and number information of the occupied channels are obtained according to the elements in the OSS.Compared with the classical BMSS methods based on the information theoretic criteria(ITC),the new method shows more excellent performance especially in the low signal-to-noise ratio(SNR)and the small sampling number scenarios,and more robust detection performance in noise uncertainty or unequal noise variance applications.Meanwhile,the new method performs more stablely than the ITC-based methods when the occupied subband number increases or the primary signals suffer multi-path fading.Simulation result verifies the effectiveness of the proposed method.

Propagation simulation and dilatancy analysis of rock joint using displacement discontinuity method

A revised displacement discontinuity method(DDM) program is developed for the simulation of rock joint propagation and dilatancy analysis. The non-linear joint model used in the program adopts Barton-Bandis normal deformation model, Kulhaway shear deformation model and Mohr-Coulomb criterion. The joint propagation criterion is based on the equivalent stress intensity factor which can be obtained by regression analysis. The simulated rock joint propagation accords well with the existing knowledge. The closure and opening of joint is investigated by DDM, and it is shown that if the opening volume of propagated joint is larger than closure volume of the old joint, the joint dilatancy occurs. The dilatancy condition is mainly controlled by the normal stiffness of the rock joint. When the normal stiffness is larger than the critical value, joint dilatancy occurs. The critical normal stiffness of rock joint changes with the joint-load angle, and joint dilatancy is most possible to occur at 30°.

Existence of Tensor Glueball ξ（2230） and Its Mass, Spin, and Decay Width

In order to solve the current experimental controversy on the existence of the tensor glueball ξ(2230), we examine whether the conclusion from Seth's experimental measurement, which rejects existence of ξ(2230), is reliable.We claim that the non-observation of the resonance structure by Seth is not a sound ground for rejecting the existence of ξ(2230) but it may just indicate that ξ(2230) has a broad width. We also study the tensor glueball properties: the mass, spin, and decay width. Our theoretical predictions of the mass and the spin of the ξ(2230) using QCD sum rule are about 2.23 GeV and J = 2, respectively. Our theoretical results also evidently show that the ξ(2230) must have a total decay width of about 100 MeV, which is much broader than the reported 20 MeV but a small partial decay width of 2 MeV into pp decay channel.

An approach for complex activity recognition by key frames

A new method for complex activity recognition in videos by key frames was presented. The progressive bisection strategy(PBS) was employed to divide the complex activity into a series of simple activities and the key frames representing the simple activities were extracted by the self-splitting competitive learning(SSCL) algorithm. A new similarity criterion of complex activities was defined. Besides the regular visual factor, the order factor and the interference factor measuring the timing matching relationship of the simple activities and the discontinuous matching relationship of the simple activities respectively were considered. On these bases, the complex human activity recognition could be achieved by calculating their similarities. The recognition error was reduced compared with other methods when ignoring the recognition of simple activities. The proposed method was tested and evaluated on the self-built broadcast gymnastic database and the dancing database. The experimental results prove the superior efficiency.