ID 原文 译文
17735 为了使分类器能够解决非线性数据分类的问题，该文通过核技巧对SMLR进行核化扩充后得到了核稀疏多元逻辑回归(KSMLR)。 In order to solve the problem of non-linear data classification, Kernel Sparse Multinomial Logistic Regression(KSMLR) is obtained by kernel trick.
17736 KSMLR能够将非线性特征数据通过核函数映射到高维甚至无穷维的特征空间中，使其特征能够充分地表达并最终能进行有效的分类。 KSMLR can map nonlinear feature data into high-dimensional and eveninfinite-dimensional feature spaces through kernel functions, so that its features can be fully expressed andeventually classified effectively.
17737 此外，该文还利用了基于中心对齐的多核学习算法，通过不同的核函数对数据进行不同维度的映射，并用中心对齐相似度来灵活地选取多核学习权重系数，使得分类器具有更好的泛化能力。 In addition, the multi-kernel learning algorithm based on centered alignment isused to map the data in different dimensions through different kernel functions. Then center-aligned similarity can be used to select flexibly multi-kernel learning weight coefficients, so that the classifier has better generalization ability.
17738 实验结果表明，该文提出的基于中心对齐多核学习的稀疏多元逻辑回归算法在分类的准确率指标上都优于目前常规的分类算法。 The experimental results show that the sparse multinomial logistic regression algorithm based on center-aligned multi-kernel learning is superior to the conventional classification algorithm in classification accuracy.
17739 针对现有L型阵列相干信号DOA估计算法精度不高、孔径损失较大的问题，该文提出一种基于主奇异矢量的解相干(L-PUMA)方法以及改进的主奇异矢量法(L-MPUMA)。 In order to handle the problem that the existing DOA estimation algorithm for L-shaped array ofcoherent signals is not accurate and the aperture loss is large, a method named L-shaped array Principal-singular-vector Utilization for Modal Analysis (L-PUMA) and its modified algorithm named L-shaped arrayModified PUMA (L-MPUMA) are proposed.
17740 L-PUMA算法首先对互协方差矩阵进行降噪，再通过奇异值分解得到2维主奇异矢量，然后利用加权最小二乘法得到线性预测方程的多项式系数，该线性预测方程的根即为信号的DOA估计，最后提出一种新的配对算法实现仰角和方位角的配对。 L-PUMA algorithm first denoises the cross-covariance matrix,then obtains the two-dimensional main singular vector by singular value decomposition, and then obtains thepolynomial coefficient of the linear prediction equation by weighted least squares method. The root of the linearprediction equation is the DOA estimation of the signals. Finally, a new pairing algorithm is proposed to realizethe pairing of elevation and azimuth.
17741 L-MPUMA算法利用反向共轭变换构造增广主奇异矢量，进一步提高了数据利用率，克服了信号完全相干时L-PUMA算法性能下降严重的问题，仿真实验验证了所提算法的高效性。 L-MPUMA algorithm uses the inverse conjugate transform to obtain the augmented main singular vector, which further improves the data utilization rate and overcomes the problem that the performance of L-PUMA deteriorates seriously when the signals are completely coherent. Simulationexperiments verify the efficiency of the proposed algorithm.
17742 现有的增广状态-交互式多模型算法存在着依赖于量测噪声协方差矩阵这一先验信息的问题。 The existing Augmented State-Interracting Multiple Model (AS-IMM) algorithm suffers from theproblem that it relies on the prior information of the covariance matrix of the measurement noise.
17743 当先验信息未知或不准确时，算法的跟踪性能将会下降。 When the prior information is unavailable or inaccurate, the tracking performance of AS-IMM will be degraded.
17744 针对上述问题，该文提出一种自适应的变分贝叶斯增广状态-交互式多模型算法VB-AS-IMM。 In orderto overcome this problem, a novel adaptive Bayesian Variational Augmented State-Interracting Multiple Model(VB-AS-IMM) algorithm is proposed.