Order and Disorder in Product Innovation Models

This article argues that the conceptual development of product innovation models goes hand in hand with paradigmatic changes in the field of organization science. Remarkable similarities in the change of organizational perspectives and product innovation models are noticeable. To illustrate how changes in the organizational paradigm are being translated into changes in new product development (NPD) practices, five NPD models are presented: the sequential, compression, flexible, integrative and improvisational models. The evolution of product innovation management shows a move from planned and mechanistic, towards emergent and organic models. Such a process of re‐orientation poses several challenges that are presented in the form of six propositions: from universal to contingent models, from invariant to flexible practices, from avoiding risks to taking advantage of opportunities, from planning to learning, from exclusive teams to inclusive networks, from structure to structured chaos.     

Radial data envelopment analysis models with mixed orientation of input and output

This paper discusses parametric solutions and envelopment formulations of radial data envelopment analysis (DEA) models with mixed orientation of input and output. These solutions geometrically but not numerically lie between the two usual solutions from input and output orientations. The consequent results provide alternative optimal solutions between those from input‐ and output‐oriented CCR models for constant returns to scale DEA models and optimal scale efficiency in addition to technical efficient solutions from input‐ and output‐oriented BCC models for variable returns to scale DEA models.     

一种网格融合算法

快速建模是三维游戏动画领域的重要技术,从已有模型经过修改、编辑、融合构建出新的模型是一种高效的建模方法.本文提出了一种网格模型融合算法,该算法首先将需要的部分网格从源模型上交互剪切下来,并将其配准对齐;然后将两网格模型转化成点模型表示,并将点模型转化成RBF隐函数表示;再对两隐函数进行布尔运算;最后将布尔运算生的隐函数曲面在两网格接合区域进行三角形化,得到最终的网格模型.算法定义了隐函数曲面的影响区域,有效控制融合过渡.采用边界扩展的三角形化方法,保留了融合区域以外源模型的特征.实验结果表明,本文算法具有很好的网格融合结果,可用于游戏动画中快速造型.     

面向边缘智能的两阶段对抗知识迁移方法

对抗样本的出现,对深度学习的鲁棒性提出了挑战.随着边缘智能的兴起,如何在计算资源有限的边缘设备上部署鲁棒的精简深度学习模型,是一个有待解决的问题.由于精简模型无法通过常规的对抗训练获得良好的鲁棒性,提出两阶段对抗知识迁移的方法,先将对抗知识从数据向模型迁移,然后将复杂模型获得的对抗知识向精简模型迁移.对抗知识以对抗样本的数据形式蕴含,或以模型决策边界的形式蕴含.具体而言,利用云平台上的GPU集群对复杂模型进行对抗训练,实现对抗知识从数据向模型迁移;利用改进的蒸馏技术将对抗知识进一步从复杂模型向精简模型的迁移,最后提升边缘设备上精简模型的鲁棒性.在MNIST,CIFAR-10和CIFAR-100这3个数据集上进行验证,实验结果表明：提出的这种两阶段对抗知识迁移方法可以有效地提升精简模型的性能和鲁棒性,同时加快训练过程的收敛性.     

Mental causal models of incidents communicated in licensee event reports in a process industry

The present investigation describes some mental causal models used in incident reports. Some of the models (e.g., single-cause models) are simpler than others (e.g., causal-tree models). The models are also associated with different ways of explaining an incident or accident and with different recommendations for increasing the safety of a system. In study 1, incident reports from Swedish nuclear power plants known to use human or organisational factors were analysed. The analysis showed that the most frequent model was a simple single-cause model. Two-step models and more complex models were less frequent. Study 2 analysed all licensee event reports (including those reports not related to human organisational factors) from four reactors assessed by regulators during the year. The results showed that single-cause and two-step accident models were more frequent than more complex models. The analyses also revealed that different detection modes were related to different models.     

Modeling reuse across the software life cycle

This article reports strong correlations between reuse levels of software products across the life cycle and presents models of reuse levels across the life cycle. These models can be used to estimate reuse levels of one life cycle object given reuse levels of others, or to predict reuse levels of later life cycle objects given reuse levels of earlier ones. The models have been developed from survey data collected from 113 respondents from 29 organizations about their own and their organizations' levels of reuse of life cycle objects. Usage scenarios for the models are presented, as are methods for customizing and improving the models.     

Do Markov-switching models capture nonlinearities in the data?: Tests using nonparametric methods

Markov-switching models have become popular alternatives to linear autoregressive models. Many papers which estimate nonlinear models make little attempt to demonstrate whether the nonlinearities they capture are of interest or if the models differ substantially from the linear option. By simulating the models and nonparametrically estimating functions of the simulated data, we can evaluate if and how the nonlinear and linear models differ.     

Querying business process model repositories

Business process management technology is becoming increasingly popular, resulting in more and more business process models being created. Hence, there is a need for these business process models to be managed effectively. For effective business process model management, being able to efficiently query large amount of business process models is essential. For example, it is preferable to find a similar or related model to customize, rather than building a new one from scratch. This would not only save time, but would also be less error-prone and more coherent with the existing models of the enterprise. Querying large amounts of business process models efficiently is also vital during company amalgamation, in which business process models from multiple companies need to be examined and integrated. This paper provides: an overview of the field of querying business process models; a summary of its literature; and a list of challenges (and some potential solutions) that have yet to be addressed. In particular, we aim to compare the differences between querying business process models and general graph querying. We also discuss literature work from graph querying research that can be used when querying business process models.     

Cross versus Within-Company Cost Estimation Studies: A Systematic Review

The objective of this paper is to determine under what circumstances individual organizations would be able to rely on cross-company-based estimation models. We performed a systematic review of studies that compared predictions from cross-company models with predictions from within-company models based on analysis of project data. Ten papers compared cross-company and within-company estimation models; however, only seven presented independent results. Of those seven, three found that cross-company models were not significantly different from within-company models, and four found that cross-company models were significantly worse than within-company models. Experimental procedures used by the studies differed making it impossible to undertake formal meta-analysis of the results. The main trend distinguishing study results was that studies with small within-company data sets (i.e., $20 projects) that used leave-one-out cross validation all found that the within-company model was significantly different (better) from the cross-company model. The results of this review are inconclusive. It is clear that some organizations would be ill-served by cross-company models whereas others would benefit. Further studies are needed, but they must be independent (i.e., based on different data bases or at least different single company data sets) and should address specific hypotheses concerning the conditions that would favor cross-company or within-company models. In addition, experimenters need to standardize their experimental procedures to enable formal meta-analysis, and recommendations are made in Section 3.     

Developing empirical models from observational data using artificial neural networks

Computer-integrated manufacturing requires models of manufacturing processes to be implemented on the computer. Process models are required for designing adaptive control systems and selecting optimal parameters during process planning. Mechanistic models developed from the principles of machining science are useful for implementing on a computer. However, in spite of the progress being made in mechanistic process modeling, accurate models are not yet available for many manufacturing processes. Empirical models derived from experimental data still play a major role in manufacturing process modeling. Generally, statistical regression techniques are used for developing such models. However, these techniques suffer from several disadvantages. The structure (the significant terms) of the regression model needs to be decided a priori. These techniques cannot be used for incrementally improving models as new data becomes available. This limitation is particularly crucial in light of the advances in sensor technology that allow economical on-line collection of manufacturing data. In this paper, we explore the use of artificial neural networks (ANN) for developing empirical models from experimental data for a machining process. These models are compared with polynomial regression models to assess the applicability of ANN as a model-building tool for computer-integrated manufacturing.Operated for the United States Department of Energy under contract No. DE-AC04-76-DP00613.     

Gaussian mixture discriminant analysis and sub-pixel land cover characterization in remote sensing

Mixture analysis is a necessary component for capturing sub-pixel heterogeneity in the characterization of land cover from remotely sensed images. Mixture analysis approaches in remote sensing vary from conventional linear mixture models to nonlinear neural network mixture models. Linear mixture models are fairly simple and generally result in poor mixture analysis accuracy. Neural network models can achieve much higher accuracy, but typically lack interpretability. In this paper we present a mixture discriminant analysis (MDA) model for inferring land cover fractions within forest stands from Landsat Thematic Mapper images. Specifically, individual class distributions are modeled as mixtures of subclasses of Gaussian distributions, and land cover fractions are estimated using the corresponding posterior probabilities. Compared to a benchmark study on accuracy of mixture models with Plumas National Forest data, this MDA model easily outperforms traditional linear mixture models and parallels the performance of the ARTMAP neural network mixture model. In other words, the MDA model is observed to successfully combine the performance characteristics of more complex neural network models (due to the nonlinear nature of its classification rules), with the ease of interpretation associated with linear mixture models (due to its relatively simple structure). MDA models therefore offer an attractive alternative for addressing the mixture modeling problem in remote sensing.     

信息检索的概率模型

The study of mathematical models on information retrieval is an important area in the Information Retrieval community. Because of the uncertainty characteristic of IR,the probability model based on statistical probability is apromising model from recent to future. Those models can be classified into classical models and probability network models. Several famous models are introduced and their shortcomings are pointed out in this paper. We also clarifythe relationship of these models and introduce a new models based on statistical language model curtly.     

A uniform approach for access control and business models with explicit rule realization

Access control is an important part of security in software, such as business applications, since it determines the access of users to objects and operations and the constraints of this access. Business and access control models are expressed using different representations. In addition, access control rules are not generally defined explicitly from access control models. Even though the business model and access control model are two separate modeling abstractions, they are inter-connected as access control is part of any business model. Therefore, the first goal is to add access control models to business models using the same fundamental building blocks. The second goal is to use these models and define general access control rules explicitly from these models so that the connection between models and their realizations are also present. This paper describes a new common representation for business models and classes of access control models based on the Resource–Event–Agent (REA) modeling approach to business models. In addition, the connection between models and their represented rules is clearly defined. We present a uniform approach to business and access control models. First, access control primitives are mapped onto REA-based access control patterns. Then, REA-based access control patterns are combined to define access control models. Based on these models, general access control rules are expressed in Extended Backus–Naur Form.     

Knowledge-based versus data-driven fuzzy habitat suitability models for river management

Aquatic habitat suitability models have increasingly received attention due to their wide management applications. Ecological expert knowledge has been frequently incorporated in such models to link environmental conditions to the quantitative habitat suitability of aquatic species. Since the formalisation of problem-specific human expert knowledge is often difficult and tedious, data-driven machine learning techniques may be helpful to extract knowledge from ecological datasets. In this paper, both expert knowledge-based and data-driven fuzzy habitat suitability models were developed and the performance of these models was compared. For the data-driven models, a hill-climbing optimisation algorithm was applied to derive ecological knowledge from the available data. Based on the available ecological expert knowledge and on biological samples from the Zwalm river basin (Belgium), habitat suitability models were generated for the mayfly Baetis rhodani (Pictet 1843). Data-driven models appeared to outperform expert knowledge-based models substantially, while a step-forward model selection procedure indicated that physical habitat variables adequately described the mayfly habitat suitability in the studied area. This study has important implications on the application of expert knowledge in ecological studies, especially if this knowledge is extrapolated to other areas. The results suggest that data-driven models can complement expert knowledge-based approaches and hence improve model reliability.     

面向自然语言处理任务的预训练模型综述

近年来,深度学习技术得到了快速发展。在自然语言处理（NLP）任务中,随着文本表征技术从词级上升到了文档级,利用大规模语料库进行无监督预训练的方式已被证明能够有效提高模型在下游任务中的性能。首先,根据文本特征提取技术的发展,从词级和文档级对典型的模型进行了分析;其次,从预训练目标任务和下游应用两个阶段,分析了当前预训练模型的研究现状,并对代表性的模型特点进行了梳理和归纳;最后,总结了当前预训练模型发展所面临的主要挑战并提出了对未来的展望。     

Bayesian non-linear regression models with skew-elliptical errors: Applications to the classification of longitudinal profiles

Typically, the fundamental assumption in non-linear regression models is the normality of the errors. Even though this model offers great flexibility for modeling these effects, it suffers from the same lack of robustness against departures from distributional assumptions as other statistical models based on the Gaussian distribution. It is of practical interest, therefore, to study non-linear models which are less sensitive to departures from normality, as well as related assumptions. Thus the current methods proposed for linear regression models need to be extended to non-linear regression models. This paper discusses non-linear regression models for longitudinal data with errors that follow a skew-elliptical distribution. Additionally, we discuss Bayesian statistical methods for the classification of observations into two or more groups based on skew-models for non-linear longitudinal profiles. Parameter estimation for a discriminant model that classifies individuals into distinct predefined groups or populations uses appropriate posterior simulation schemes. The methods are illustrated with data from a study involving 173 pregnant women. The main objective in this study is to predict normal versus abnormal pregnancy outcomes from beta human chorionic gonadotropin data available at early stages of pregnancy.     

System-on-chip design by proof-based refinement

Systems-on-chip (SoCs) and SoC architectures provide a collection of challenging problems related to specification, modelling techniques, security issues and structuring questions. We describe a design methodology integrating the event B method and characterized by the incremental and proof-controlled construction of SoC models. The essence of the methodology is the refinement of models, starting from system requirements and producing event B models for characterizing the system under development. The refinement is a unifying concept that ensures the consistency of the different models produced and our contribution is an illustration through a case study, namely a system for measuring the parameters of audio/video quality in the digital video broadcasting (DVB) set of digital TV standards. The first part is the derivation of an architecture of parameters from the document ETSI TR 101 290 and the validation of the architecture using invariants of B models. The second part is the proposal of B models of the SystemC scheduler and an instantiation of these abstract models of the simulation semantics by parameters of the SystemC codes automatically translated from the B models of the DVB system. Finally, the third part relies upon a proof-based methodology for deriving an operational semantics of a given system that is expressed by an event B model including invariant properties.     

Development life cycle of computer networks: the executable modelapproach

An approach is proposed for extending the use of design models to the implementation and operational phases of the network development life cycle. A conceptual approach is offered for using executable models in the day-to-day operation of computer networks. Several strategies are given for integrating models into different development tasks. It is shown why these strategies are feasible. Characteristics of executable models that differ from those of traditional models are identified along with new technologies that reduce the cost of implementing and using executable models     

Support vector regression for anomaly detection from measurement histories

This research focuses on the analysis of measurements from distributed sensing of structures. The premise is that ambient temperature variations, and hence the temperature distribution across the structure, have a strong correlation with structural response and that this relationship could be exploited for anomaly detection. Specifically, this research first investigates whether support vector regression (SVR) models could be trained to capture the relationship between distributed temperature and response measurements and subsequently, if these models could be employed in an approach for anomaly detection. The study develops a methodology to generate SVR models that predict the thermal response of bridges from distributed temperature measurements, and evaluates its performance on measurement histories simulated using numerical models of a bridge girder. The potential use of these SVR models for damage detection is then studied by comparing their strain predictions with measurements collected from simulations of the bridge girder in damaged condition. Results show that SVR models that predict structural response from distributed temperature measurements could form the basis for a reliable anomaly detection methodology.