ID 原文 译文
11374 如何构建基于数据的编码矩阵是编码方法确定的关键。 ‭How to build based on the data encoding matrix is the key to the encoding method to determine.
11375 针对此问题,基于Fisher原理,提出一种基于支持向量数据描述(support vector domain description,SVDD)的层次纠错输出编码构造方法(hierarchical error-correcting output codes,HECOC)。 ‭Aiming at this problem, based on Fisher principle, this paper puts forward a kind of based on support vector data description (the support vector domain description, the SVDD) the level of output error correction coding method (hierarchical error - correcting output codes, HECOC).
11376 该方法首先采用SVDD计算各类别的可分程度,从而得到由不同子类构成的二叉树;‭ ‭This method firstly all kinds of other can be divided into degree, is obtained by using the SVDD is obtained by different subclasses of binary tree; ‭
11377 然后分别对二叉树的各层结点进行编码并最终形成层次输出编码。 Then respectively to encode binary tree nodes of every layer and finally form the level of output encoding.
11378 在仿真实验中,对不同子类类群划分构成的基分类器的可分性进行了对比。 ‭In simulation experiments, the different living groups constitute the base classifier divided into subclasses of separability are compared.
11379 结果表明,该编码方法能在保证分类精度的同时,提高基分类器之间的差异性和纠错输出编码的容错能力。 The results show that the coding method can ensure the classification accuracy at the same time, improve the difference between the base classifier and the tolerance of error correcting output codes.
11380 提出了一种新型的基于Hammerstein-Wiener模型的广义预测控制策略。 Put forward a new kind of generalized predictive control strategy based on Hammerstein - Wiener model.
11381 采用基于最小二乘支持向量机的Hammerstein-Wiener模型描述非线性系统动态特性,作为被控对象预测模型。 Based on least squares support vector machine (SVM) Hammerstein - Wiener model describe the nonlinear system dynamic characteristics, as a prediction model of the object.
11382 同时,针对现有遗传算法和混沌粒子群优化算法收敛速度慢和精度低等缺点,给出一种拟牛顿信赖域混沌粒子群混合优化算法。 At the same time, in view of the existing genetic algorithm and chaotic particle swarm optimization algorithm is slow convergence speed and low accuracy of shortcomings, given a quasi-newton trust region chaos particle swarm hybrid optimization algorithm.
11383 作为预测控制的滚动优化策略,函数测试和非线性对象的广义预测控制的滚动优化表明该算法的优越性。 As the rolling optimization strategy of predictive control, function test and generalized predictive control of nonlinear object rolling optimization show that the superiority of the algorithm.