The Morris elementary effects method (MM) is a widely used, model-free approach for factor screening and sensitivity analysis across various domains. However, traditional MM can be computationally demanding due to its "one-factor-at-a-time" nature. This paper presents a novel multi-stage sequential-group Morris method (SGMM) and a corresponding sequential implementation procedure tailored for both deterministic and stochastic simulation settings. SGMM leverages prior knowledge of the directional influence of individual factors to define elementary effects at the group level. This enables early elimination of factor groups with negligible group effects. Each stage employs a distribution-free sequential probability ratio test (SPRT) to evaluate the significance of group effects, ensuring rigorous control over Type I and Type II familywise error rates. Numerical experiments show that SGMM consistently outperforms existing simulation-based factor screening methods, delivering up to 87.87% computational savings while preserving high statistical accuracy.

Quality and safety are critical for manufacturers, particularly in the automotive industry, where rising defect complaints drive vehicle recalls. The growth of platform economics and social media has increased complaint volumes, posing challenges for government agencies and companies in managing them effectively. This work introduces a robust analytics approach to assess and optimise the complaint-recall process, featuring two main components: agent-based modelling and multi-response simulation factor screening. The agent-based simulation model serves as an experimental testbed, revealing complex behaviours and dynamics within the complaint-recall process. An extended version of Morris' elementary effects method is proposed to identify key operational factors impacting performance. The numerical study demonstrates that sustained increases in defect complaints can significantly degrade system performance, particularly in terms of delays. Our findings suggest that both government agencies and manufacturers play crucial roles in managing flow times for recall and non-recall cases. Enhanced cooperation between these parties can notably reduce recall case processing times. This study provides stakeholders with a valuable tool for optimising operational conditions and improving overall recall system efficiency.

Estimating the maximum mean finds a variety of applications in practice. In this paper, we study estimation of the maximum mean using an upper confidence bound (UCB) approach where the sampling budget is adaptively allocated to one of the systems. We study in depth the existing grand average (GA) estimator, and propose a new largest-size average (LSA) estimator. Specifically, we establish statistical guarantees, including strong consistency, asymptotic mean squared errors, and central limit theorems (CLTs) for both estimators, which are new to the literature. We show that LSA is preferable over GA, as the bias of the former decays at a rate much faster than that of the latter when sample size increases. By using the CLTs, we further construct asymptotically valid confidence intervals for the maximum mean, and propose a single hypothesis test for a multiple comparison problem with application to clinical trials. Statistical efficiency of the resulting point and interval estimates and the proposed single hypothesis test is demonstrated via numerical examples.

Morris' elementary effect-based screening (MM) has been widely used in a variety of domains to identify a few important factors among many possible ones. In MM, cluster sampling offers substantial computational savings over non-cluster sampling, but it remains a challenge to construct a cluster sampling matrix that generates any particular number of elementary effects for each factor. In this paper, we thoroughly address this issue. We uncover the mathematical association between distinct block sampling matrices within the complete cluster sampling matrix, by introducing a "dummy" sub-block matrix. By leveraging this property, we propose an easy-to-implement adaptive cluster sampling (ACS) method that is capable of identifying the appropriate sub-block sampling matrix to use. Its advantage over existing brute-force methods is that it can provide easy-to-obtain cluster sampling matrices, and it can be applied to computational model with any number of factors given a prior. We demonstrate the attractive properties of ACS using analytical proofs and simulation experiments. We show the robustness of ACS via a real-world case study. Our code for the algorithm is available online. (c) 2023 Elsevier B.V. All rights reserved.

In the context of digital economy, businesses are increasingly involved in diverse digital innovation initiatives with the aim of achieving high product quality. Despite this, there is no unanimous consensus on product quality, even in the realm of traditional technological innovation. Therefore, it is worth contemplating whether digital innovation in business invariably leads to enhanced product quality. This study investigates the interplay between the scale and value of digital innovation (DI-S and DI-V) and product recalls, which serve as a crucial indicator of product quality, while considering product variety as a moderating factor. The analysis is based on 1,408 digital patent data collected from 31 automotive companies engaged in digital innovation activities from 2012 to 2020 in China. The findings reveal that while DI-S does not yield the expected significant impact on product quality, DI-V has a positive influence on product quality. Furthermore, businesses with a broader range of product categories can enhance product quality more effectively by focusing on improving DI-V, especially when compared to those with fewer product categories. This research offers practical insights into the implementation of digital innovation by businesses and holds significant importance in motivating government actions aimed at enhancing policies for achieving high-quality development.

Product recalls are highly disruptive for many firms. Understanding the drivers of such recalls is paramount to helping firms effectively reduce product recall risk. While prior studies have investigated the drivers of product recalls in developed markets, little is known about the factors that drive product recalls in emerging markets. Using data for 2010–2016, this study identifies firm innovation and negative electronic word-of-mouth (eWOM) as drivers that influence the volume of vehicle recalls in the Chinese automobile industry, a sector characterized by increasing R&D investment and consumer quality awareness. Considering the foreign ownership restriction policy of the Chinese automobile industry, we further examine the moderating effect of ownership structure. We find that firm innovation increases the volume of product recalls for Chinese domestic automakers (CDAs) while decreasing the volume of product recalls for international joint ventures (IJVs), negative eWOM leads automakers to recall defective vehicles, and the ownership structure itself (IJV versus CDA) has a significant impact on the volume of product recalls. These results offer insights that can help managers take concrete steps to reduce and counter product recalls.

We propose a novel screening framework (abbreviated to TopmSB) to identify the top m key factors affecting the system performance. The new framework extends the standard SB’s screening mechanism while in each stage marrying with an adaptive multi-armed bandit (MAB) procedure in order of the largest group. Compared to SB, TopmSB avoids specifying perplexing (un)importance threshold parameters, while providing desired computational efficiency and statistical precision guarantee. Numerical experiments demonstrate the efficiency and effectiveness of the proposed method.

辨证分型是中医诊断治疗中的一个重要环节,首先要诊断患者疾病,提出相应的辨证方法,并进行证候分型以指导治疗,同时证型可以从古代方剂中归纳和总结.文章采用遗传算法原理对古代闭经方剂进行聚类,得到的方剂聚类簇即可作为证型总结和归纳的依据,为当代中医诊断智能化提供思路.

Syndrome types are important for diagnosis and treatment in traditional Chinese medicine. Syndrome types can be summarized by domain experts as formula clusters. In this paper, we propose seven feature models for the formula clustering problem based on categories, subcategories, functional tendencies and names of Chinese materia medica. A novel multi-objective clustering hyper-heuristic algorithm is obtained. In our proposed algorithm, 12 low-level heuristics are used for clustering solution perturbation by merging clusters, dividing clusters or moving points between clusters based on received solutions from the high-level heuristic. The high-level heuristic evaluates the received solutions from low-level heuristics, updates the solution pool, and selects initial solutions for the next iteration via roulette wheel selection on the Pareto front. Experimental results demonstrate that the proposed algorithm outperforms other clustering algorithms in most datasets. The initial number of clusters has less influence on the final clustering solutions for our proposed algorithm than for other clustering algorithms. For most datasets, the roulette wheel selection mechanism on the Pareto front shows higher convergence rates and accuracy than a random selection mechanism. Accuracy was higher for feature models based on functional tendencies than for the other feature models.

The analysis of engineering systems via computer-based models often confronts the intricacy of high-dimensional input spaces and notable interrelationships between inputs. This, in turn, necessitates the implementation of sensitivity analysis as a means of guiding reliability engineering. However, the query of effectively identifying significant effects, particularly interaction effects, remains unresolved. In this paper, we present a two-stage elementary effect-based sensitivity analysis method that effectively identifies main and interaction effects by fully utilizing the sequential characteristics of simulation experiments. Not only can the proposed method efficiently identify factors with important (a) main effects and (b) interaction and/or non-linear effects, but more importantly, it can discern the important two-factor interaction effects. Compared to state-of-the-art elementary effect-based methods, the proposed procedure can additionally identify specific interaction effects between two factors. Compared to traditional second-order elementary effect-based method, our method can achieve enormous computational savings without sacrificing statistical effectiveness. The Monte Carlo simulation experiments verify the feasibility and the real-world case study conducted to design a reliable cross-docking center manifests the robustness of the proposed method.

Syndrome is a crucial principle of Traditional Chinese Medicine. Formula classification is an effective approach to discover herb combinations for the clinical treatment of syndromes. In this study, a local search based firefly algorithm (LSFA) for parameter optimization and feature selection of support vector machines (SVMs) for formula classification is proposed. Parameters C and gamma of SVMs are optimized by LSFA. Meanwhile, the effectiveness of herbs in formula classification is adopted as a feature. LSFA searches for well-performing subsets of features to maximize classification accuracy. In LSFA, a local search of fireflies is developed to improve FA. Simulations demonstrate that the proposed LSFA-SVM algorithm outperforms other classification algorithms on different datasets. Parameters C and gamma and the features are optimized by LSFA to obtain better classification performance. The performance of FA is enhanced by the proposed local search mechanism.

In recent years, the problems of low degree of industrialization of agriculture, weak informatization ability and food safety have become increasingly serious. This article combines the detection of agricultural products supply chain and RFID technology and applies it to the testing of agricultural products supply chain. In this study, the agricultural product supply chain and agricultural product logistics information system were introduced. At the same time, the application of RFID technology in the production, processing and other aspects of the detection of agricultural products supply chain is elaborated, and the information system of RFID technology in the agricultural product supply chain is designed. Finally, the efficiency of RFID technology in the detection of agricultural products supply chain has been verified. Therefore, this technology is the future trend of agricultural logistics development, thereby promoting the development of agricultural products logistics supply chain testing.

In this paper we provide a thorough investigation of the cluster sampling scheme for Morris' elementary effects method (MM), a popular model‐free factor screening method originated in the setting of design and analysis of computational experiments. We first study the sampling mechanism underpinning the two sampling schemes of MM (i.e., cluster sampling and noncluster sampling) and unveil its nature as a two‐level nested sampling process. This in‐depth understanding sets up a foundation for tackling two important aspects of cluster sampling: budget allocation and sampling plan. On the one hand, we study the budget allocation problem for cluster sampling under the analysis of variance framework and derive optimal budget allocations for efficient estimation of the importance measures. On the other hand, we devise an efficient cluster sampling algorithm with two variants to achieve enhanced statistical properties. The numerical evaluations demonstrate the superiority of the proposed cluster sampling algorithm and the budget allocations derived (when used both separately and in conjunction) to existing cluster and noncluster sampling schemes.

中医方剂配伍规律是指不同的中药组合会呈现多样的治疗效果,针对治疗同一类疾病的方剂往往具有某些类似的中药配伍.笔者以改进遗传算法为基础,以关联规则中的置信度和支持度设计适应度函数,同时以古代胸痹心痛方剂件为初始化种群进行迭代,挖掘出针对胸痹心痛病的常用药物配伍,旨在为当代中医药方剂治疗胸痹心痛提供思路.

Because computers (except for parallel computers) generate simulation outputs sequentially, we recommend sequential probability ratio tests (SPRTs) for the statistical analysis of these outputs. However, until now simulation analysts have ignored SPRTs. To change this situation, we review SPRTs for the simplest case; namely, the case of choosing between two hypothesized values for the mean simulation output. For this case, the classic SPRT of Wald (Wald A. Sequential tests of statistical hypotheses. Ann Math Stat 1945; 16: 117–186) allows general types of distribution, including normal distributions with known variances. A modification permits unknown variances that are estimated. Hall (Hall WJ. Some sequential analogs of Stein’s two-stage test. Biometrika 1962; 49: 367–378) developed a SPRT that assumes normal distributions with unknown variances estimated from a pilot sample. A modification uses a fully sequential variance estimator. In this paper, we quantify the performance of the various SPRTs, using several Monte Carlo experiments. In experiment #1, simulation outputs are normal. Whereas Wald’s SPRT with estimated variance gives too high error rates, Hall’s original and modified SPRTs are “conservative”; that is, the actual error rates are smaller than those prespecified (nominal). Furthermore, our experiment shows that the most efficient SPRT is Hall’s modified SPRT. In experiment #2, we estimate the robustness of these SPRTs for non-normal output. For these two experiments, we provide details on their design and analysis; these details may also be useful for simulation experiments in general.

关于中医方剂的古籍资料浩如烟海,但历代医家对于方剂的分类却始终未能形成一套统一的规则.为了实现对众多方剂的分类规则挖掘,需要通过计算机构造合适的分类器实现.笔者运用决策树算法对213例古代胸痹心痛方剂进行识别和分类,得到正确率较高的分类结果,为中医方剂分类研究领域提供参考和借鉴.

针对在复杂情景下视频前背景分离技术中存在的前景泄露问题,设计开发了一个端对端的二级级联深度卷积神经网络,实现了对输入视频序列进行精确的前景和背景分离.所提网络由一级前景检测子网络和二级背景重建子网络串联而成.一级网络融合时间和空间信息,其输入包含2个部分:第1个部分是3张连续的彩色RGB视频帧,分别为上一帧、当前帧和下一帧;第2个部分是3张与彩色视频帧相对应的光流图.一级前景检测子网络通过结合2部分输入对视频序列中运动的前景进行精确检测,生成二值化的前景掩膜.该部分网络是一个编码器-解码器网络:编码器采用VGG16的前5个卷积块,用来提取两部分输入的特征图,并在经过每一个卷积层后对两类特征图进行特征融合;解码器由5个反卷积模块构成,通过学习特征空间到图像空间的映射,从而生成当前帧的二值化的前景掩膜.二级网络包含3个部分:编码器、传输层和解码器.二级网络能够利用当前帧和生成的前景掩膜对缺失的背景图像进行高质量的修复重建.实验结果表明,本文所提时空感知级联卷积神经网络在公共数据集上取得了较其他方法更好的结果,能够应对各种复杂场景,具有较强的通用性和泛化能力,且前景检测和背景重建结果显著超越多种现有方法.

The past decade has witnessed a remarkable growth of automobile sales and production in emerging economies, with China developing into the largest global auto market since 2009. This paper focuses on an important but neglected aspect in these emerging markets, namely, vehicle recalls. The aim of this study is twofold. The first is to show that a significant difference exists in the number and volume of vehicle recalls between the emerging Chinese market and the established US market; the second is to detect whether this difference can be attributed to the initiator level (voluntary versus involuntary recalls) and/or the firm level (organizational ownership structure and nationality of the foreign partner in international joint ventures). To that end, we quantify the recall performance by means of 4 metrics: total number of recall events per annum (NRE), total number of units recalled per annum (NUR), average number of vehicles recalled per event per annum (NRPE), and recall rate (RR); for each of these, we benchmark the US market and assess the relative performance of the investigated market using a bootstrap method. The empirical results indicate that the recall metrics in the Chinese market have underperformed relative to those of the established market. This is extremely pertinent in light of the current “Going Out” policy put forward by the Chinese government, as subpar quality awareness hampers the successful access of Chinese automakers to foreign markets.