ID 原文 译文
10654 假定产品在使用时经历多种竞争失效,每种失效寿命数据服从Lindley分布。 Assume that product when use experience a variety of competitive failure, each failure life data obey Lindley distribution.
10655 竞争失效数据包括不完全数据和截尾数据。 ‭Failure data including the incomplete data and competition of according to.
10656 对竞争失效分布进行参数点估计、相对风险率计算,讨论了参数的渐进置信区间和Bootstrap置信区间。 ‭On the competition failure distribution parameter point estimation, relative risk rate calculation, discusses the parameters of the gradual confidence interval and the Bootstrap confidence intervals.
10657 为考察其适应性,一组实际数据用来说明各种参数估计方法的使用情况,同时,分别用指数分布和威布尔分布拟合该组数据,并计算相应的统计量。 ‭To inspect its adaptability, a group of actual data used to illustrate the usage of various parameter estimation methods, at the same time, respectively, with the exponential distribution and weibull distribution fitting the group data, and calculate the corresponding statistics.
10658 结果表明,根据3种分布的极大似然估计量和K-S值,Lindley分布对该组数据具有最好的适应性。 ‭Results show that according to the three kinds of distribution of the maximum likelihood estimator and K - S value, Lindley distribution data has better adaptability to the group. ‭
10659 为了充分利用稀疏表示分类信息和高光谱图像的空间信息,提出结合马尔可夫随机场的加权条件稀疏表示高光谱图像分类算法。 In order to make full use of sparse representation classification information and spatial information of hyperspectral image, put forward in combination with markov random field weighted conditional sparse representation of hyperspectral image classification algorithm.
10660 该算法对稀疏表示分解后的残差向量建立条件稀疏表示模型,在计算残差向量的类别归属时引入频段方差信息; The algorithm of sparse representation condition for establishing the residual vector of the decomposed sparse representation model, when calculating the residual vector of a nd introduced frequency variance information;
10661 利用光谱信息散度从信息熵的角度挖掘重构光谱中的类别鉴定信息; Using spectral information divergence from the Angle of information entropy, mining categories in refactoring spectrum identification information;
10662 在期望最大化算法模型中,将条件稀疏模型与光谱信息散度模型相结合,使算法具备迭代自更新的能力; In the expectation maximization algorithm model, the condition of sparse model combined with spectral information divergence model, since the update algorithm with iterative ability;
10663 将马尔可夫随机场引入加权条件稀疏表示算法,在算法时间复杂度不变的情况下,对高光谱图像的空间信息予以提取。 Markov random field to sparse representation algorithm weighted conditions, on the basis of the algorithm time complexity, the spatial information of hyperspectral image extraction.