Reliability of Space-Shuttle Pressure Vessels With Random Batch Effects

In this article we revisit the problem of estimating the joint reliability against failure by stress rupture of a group of fiber-wrapped pressure vessels used on Space-Shuttle missions. The available test data were obtained from an experiment conducted at the U.S. Department of Energy Lawrence Livermore Laboratory (LLL) in which scaled-down vessels were subjected to life testing at four accelerated levels of pressure. We estimate the reliability assuming that both the Shuttle and LLL vessels were chosen at random in a two-stage process from an infinite population with spools of fiber as the primary sampling unit. Two main objectives of this work are (1) to obtain practical estimates of reliability taking into account random spool effects and (2) to obtain a realistic assessment of estimation accuracy under the random model. Here, reliability is calculated in terms of a “system” of 22 fiber-wrapped pressure vessels, taking into account typical pressures and exposure times experienced by Shuttle vessels. Comparisons are made with previous studies. The main conclusion of this study is that, although point estimates of reliability are still in the “comfort zone,” it is advisable to plan for replacement of the pressure vessels well before the expected lifetime of 100 missions per Shuttle Orbiter. Under a random-spool model, there is simply not enough information in the LLL data to provide reasonable assurance that such replacement would not be necessary.     

Subsampling tests for the mean change point with heavy-tailed innovations

The quest of the mean change point with innovations in the domain of attraction of a κ$\kappa$-stable law appears to still be ongoing. We adopt the residual CUSUM of squares test (RCUSQ) and derive its null asymptotic distribution, which is dependent on stable index κ$\kappa$. Then a residual-based subsampling is proposed to approximate the null distribution when stable index κ$\kappa$ is unknown. Consistency and the rate of convergence for the estimated change point are also obtained. We establish the asymptotic validity of this method and assess its performance both theoretically and numerically.     

An improved universal subsampling strategy for compressing mosaic videos with arbitrary RGB color filter arrays in H.264/AVC

This work proposes an improved version of Yang et al.’s universal subsampling strategy for compressing mosaic videos with arbitrary red–green–blue (RGB) color filter arrays in H.264/AVC. For the subsampled images with 4:2:0 format, Yang et al.’s work retains the original Y luma component, but samples the proper U and V chroma components according to the corresponding mosaic structure for better reconstructing R and B pixels. However, Yang et al.’s strategy suffers from the RGB color deviations due to the U and V chroma subsampling, which results in the quality degradation of the reconstructed mosaic videos. We propose a novel modification for the Y luma component instead of only retaining the Y luma component such that the RGB color deviation problem can be resolved. Experimental results demonstrate that the proposed improved universal subsampling strategy delivers better quality of the reconstructed mosaic and full-color videos, when compared with Yang et al.’s one.     

Out-of-bag estimation of the optimal sample size in bagging

The performance of m-out-of-n bagging with and without replacement in terms of the sampling ratio (m/n) is analyzed. Standard bagging uses resampling with replacement to generate bootstrap samples of equal size as the original training set m_wor=n. Without-replacement methods typically use half samples m_wr=n/2. These choices of sampling sizes are arbitrary and need not be optimal in terms of the classification performance of the ensemble. We propose to use the out-of-bag estimates of the generalization accuracy to select a near-optimal value for the sampling ratio. Ensembles of classifiers trained on independent samples whose size is such that the out-of-bag error of the ensemble is as low as possible generally improve the performance of standard bagging and can be efficiently built.     

Design of an ADC for subsampling video applications

This paper describes a 10 bit 30 Msample/s (MSPS) CMOS analog-to-digital converter (ADC) for high-speed signal processing, especially for subsampling applications, for example digital video broadcasting over cable (DVB-C), terrestrial (DVB-T) and handheld (DVB-H) systems. The proposed pipelined ADC shows a good figure-of-merit (FoM). It adopts a power efficient amplifier sharing technique, a symmetrical gate-bootstrapping technique with modified timing for the bottom-sampling switch of a wideband sample-and-hold (S/H) circuit, a proposed stable high-swing bias circuit for a wide-swing gain-boosting telescopic amplifier. The measured differential and integral nonlinearities of the prototype in a 0.25-μm CMOS technology show less than 0.4 least significant bit (LSB) and 0.85 LSB respectively at full sampling rate. The ADC exhibits higher than 9 effective number of bits (ENOB) for input frequencies up to about 60 MHz, which is the fourfold Nyquist rate (fs/2), at 30 MSPS. The ADC consumes 60 mW from a 3-V supply and occupies about 1.36 mm². Jian Li received the Bachelor of Engineering (B.E.) degree in electronic engineering from Xi’an Jiaotong University, Xi’an, China, in 2003. He is currently working toward the Ph.D. degree at Microelectronics department, Fudan University, Shanghai, China. His current research interest is high-speed high resolution A/D converter design. Xiaoyang Zeng was born in Hunan Province, P.R. China on April 17, 1972. He received the B.S. degree from Xiangtan University, China in 1992, and the Ph.D. degree from Changchun Institute of Optics and Fine Mechanics, Chinese Academy of Sciences in 2001. From 2001 to 2003, he worked as a post-doctor researcher at the State-Key Lab of ASIC & System, Fudan University, P.R. China. Then he joined the faculty of Department of Micro-electronics at Fudan University as an associate professor. His research interests include information security chip design, VLSI signal processing, and communication systems. Prof. Zeng is the Chair of Design-Contest of ASP-DAC 2004 and 2005, also the TPC member of several international conferences such as ASCON 2005 and A-SSCC 2006, etc. Jianyun Zhang received the B.S., M.S. and Ph.D degree in electrical engineering from Fudan University, Shanghai, China in 1997, 2000 and 2006 respectively. From 2000 to 2002, he was with Alcatel microelectronics, Belgium, where he was involved in circuit design for GSM and GPRS. In 2002, he joined Trident microsystem, where he concentrated on the design of Video AFE including data converters and mixed signal circuits. In 2005, he joined Shihong microelectronics Corp., where he is now a director of mixed signal IC for video high speed interface. His research interests include data conversion, HDMI SerDes, and analog circuit design. Lei Xie received the Bachelor of Science (B.S.) degree in microelectronics from Nankai University, Tianjin, China, in 2005. He is currently working toward the M.S. degree at Fudan University, Shanghai, China. His current research interest is high-speed high resolution A/D converter. Huan Deng received the B.S. degree in microelectronics from Fudan University, Shanghai, P.R. China, in 2003. He is currently working toward the M.S. degree in microelectronics at the State Key Lab of ASIC & System, Fudan University. He is currently involved in the design of low-power, high-speed PLL’s. Yawei Guo received the B.S. and M.S. degree in electrical engineering from Fudan University in 1999 and 2002 respectively. From 2002 to August 2003, he was with Philips Semiconductors in Shanghai. Since August 2003, he has been with Shanghai MicroScience Integrated Circuits Co., Ltd., based in Shanghai, P. R. China. He has been leading a group and developing analog and mixed signal circuits. His research interests include high-speed data communication, data converters, and phase locked loops.     

Sample comminution for mycotoxin analysis: Dry milling or slurry mixing?

A comparison was made between dry milling and slurry mixing as a comminuting step preceding mycotoxin analysis. Sample schemes of up to 30 kg are mandated by European Commission legislation. Cocoa, green coffee, almonds and pistachio samples of 10 kg were milled by a Romer analytical sampling mill and all three subsamples were analysed for aflatoxin B₁ or ochratoxin A content. The homogenization process was evaluated in terms of the analytical results, coefficients of variation for different mills and particle size distributions. Coefficients of variation for the comminuting step were higher for dry milling than for slurry mixing. This difference was explained based on measured particle size distributions for both milling types. Measurements also showed slight differences in mycotoxin content of samples based on milling procedures. This might lead to lots being wrongly accepted or rejected based on an erroneous subsample result. It was concluded that sample comminution was best performed by slurry mixing, which produced smaller particles and, consequently, homogeneous samples with lowest coefficients of variation. Additional data are given on analytical results in 10-kg subsamples that originate from the aggregate 30-kg sample as described in Commission Directive 98/53/EC.     

一种大规模支持向量机的高效求解算法

现有大规模支持向量机求解算法需要大量的内存资源和训练时间,通常在大集群并行环境下才能实现。提出了一种大规模支持向量机(SVM)的高效求解算法,以在个人PC机求解大规模SVM。它包括3个步骤:首先对大规模样本进行子采样来降低数据规模;然后应用随机傅里叶映射显式地构造随机特征空间,使得可在该随机特征空间中应用线性SVM来一致逼近高斯核SVM;最后给出线性SVM在多核环境下的并行实现方法以进一步提高求解效率。标准数据集的对比实验验证了该求解算法的可行性与高效性。     

A new boosting design of Support Vector Machine classifiers

Boosting algorithms pay attention to the particular structure of the training data when learning, by means of iteratively emphasizing the importance of the training samples according to their difficulty for being correctly classified. If common kernel Support Vector Machines (SVMs) are used as basic learners to construct a Real AdaBoost ensemble, the resulting ensemble can be easily compacted into a monolithic architecture by simply combining the weights that correspond to the same kernels when they appear in different learners, avoiding to increase the operation computational effort for the above potential advantage. This way, the performance advantage that boosting provides can be obtained for monolithic SVMs, i.e., without paying in classification computational effort because many learners are needed. However, SVMs are both stable and strong, and their use for boosting requires to unstabilize and to weaken them. Yet previous attempts in this direction show a moderate success.In this paper, we propose a combination of a new and appropriately designed subsampling process and an SVM algorithm which permits sparsity control to solve the difficulties in boosting SVMs for obtaining improved performance designs. Experimental results support the effectiveness of the approach, not only in performance, but also in compactness of the resulting classifiers, as well as that combining both design ideas is needed to arrive to these advantageous designs.     

Subsampling techniques and the Jackknife methodology in the estimation of the extremal index

For a sequence of independent, identically distributed random variables any limiting point process for the time normalized exceedances of high levels is a Poisson process. However, for stationary dependent sequences, under general local and asymptotic dependence restrictions, any limiting point process for the time normalized exceedances of high levels is a compound Poisson process, i.e., there is a clustering of high exceedances, where the underlying Poisson points represent cluster positions, and the multiplicities correspond to the cluster sizes. For such classes of stationary sequences there exists the extremal indexθ, 0?θ?1, directly related to the clustering of exceedances of high values. The extremal index θ is equal to one for independent, identically distributed sequences, i.e., high exceedances appear individually, and θ>0 for “almost all” cases of interest. The estimation of the extremal index through the use of the Generalized Jackknife methodology, possibly together with the use of subsampling techniques, is performed. Case studies in the fields of environment and finance will illustrate the performance of the new extremal index estimator comparatively to the classical one.