Improved Transferability of Data‐Driven Damage Models Through Sample Selection Bias Correction

Damage models for natural hazards are used for decision making on reducing and transferring risk. The damage estimates from these models depend on many variables and their complex sometimes nonlinear relationships with the damage. In recent years, data‐driven modeling techniques have been used to capture those relationships. The available data to build such models are often limited. Therefore, in practice it is usually necessary to transfer models to a different context. In this article, we show that this implies the samples used to build the model are often not fully representative for the situation where they need to be applied on, which leads to a “sample selection bias.” In this article, we enhance data‐driven damage models by applying methods, not previously applied to damage modeling, to correct for this bias before the machine learning (ML) models are trained. We demonstrate this with case studies on flooding in Europe, and typhoon wind damage in the Philippines. Two sample selection bias correction methods from the ML literature are applied and one of these methods is also adjusted to our problem. These three methods are combined with stochastic generation of synthetic damage data. We demonstrate that for both case studies, the sample selection bias correction techniques reduce model errors, especially for the mean bias error this reduction can be larger than 30%. The novel combination with stochastic data generation seems to enhance these techniques. This shows that sample selection bias correction methods are beneficial for damage model transfer.     

论适应性表征及其方法论意义

适应性是主体适应其环境的特性,是物理演化系统和生物演化系统的本质属性,反映在认知上就是适应性表征。适应性表征也因此成为知识显现的方式和创造的核心,具有协调性、匹配性、互补性、模拟性和类比性,其表现方式有直接具象表征、间接具象表征、直接抽象表征和间接抽象表征,分别对应于经验主义、建构经验主义、理性主义和科学实在论。适应性表征的实现是通过科学理论核心概念的变化、定律的凝练、理论的更替、模型推理和世界观的改变展开的,具有经验上的适当性。这种经验适当性不仅是感知层次的体验,更是一种认知模式。在认知意义上,经验也是基于心理模型的,心理建模的路径有思想语言、心理表象、心理命题、思想实验和心理模拟推理。这种基于心理建模的适应性表征为科学认知和科学发现提供了有益的方法论。     

Perceived group cohesion versus actual social structure: A study using social network analysis of egocentric Facebook networks

Research on group cohesion often relies on individual perceptions, which may not reflect the actual social structure of groups. This study draws on social network theory to examine the relationship between observable structural group characteristics and individual perceptions of group cohesion. Leveraging Facebook data, we extracted and partitioned the social networks of 109 participants into groups using a modularity algorithm. We then surveyed perceptions of cohesion, and computed group density and size using social network analysis. Out of six linear mixed effects models specified, a random intercept and fixed slope model with group size as a predictor of perceived group cohesion emerged as best fitting. Whereas group density was not linked to perceived cohesion, size had a small negative effect on perceived cohesion, suggesting that people perceive smaller groups as more cohesive. We discuss the potential of social network analysis, visualization tools, and Facebook data for advancing research on groups.     

Setting socially mediated engagement parameters: A topic modeling and text analytic approach to examining polarized discourses on Gillette’s campaign

Opinions on social and political issues can be easily polarized in socially mediated contexts. Using an inductive topic modeling and text analytical approach, the study analyzes more than 100,000 original tweets on polarized discussions on Gillette’s controversial campaign addressing toxic masculinity. Results suggest that influencers serve as ideological hyper-prototypes that attract significant attention equally from supporters and detractors, while hashtags and users’ reference of current events span the conversations across multiple contexts. Study results also indicate that horizontal information cascades from external ideologically centered sources that serve users’ identity and ideological signaling are a potential contributor of discourse polarization. Based on the study results, an identity- and representation-based approach to understanding socially mediated discourses is proposed, and four parameters of socially mediated engagement identified to further theorize engagement for public relations research in digital contexts.     

Max‐infinitely divisible models and inference for spatial extremes

For many environmental processes, recent studies have shown that the dependence strength is decreasing when quantile levels increase. This implies that the popular max‐stable models are inadequate to capture the rate of joint tail decay, and to estimate joint extremal probabilities beyond observed levels. We here develop a more flexible modeling framework based on the class of max‐infinitely divisible processes, which extend max‐stable processes while retaining dependence properties that are natural for maxima. We propose two parametric constructions for max‐infinitely divisible models, which relax the max‐stability property but remain close to some popular max‐stable models obtained as special cases. The first model considers maxima over a finite, random number of independent observations, while the second model generalizes the spectral representation of max‐stable processes. Inference is performed using a pairwise likelihood. We illustrate the benefits of our new modeling framework on Dutch wind gust maxima calculated over different time units. Results strongly suggest that our proposed models outperform other natural models, such as the Student‐t copula process and its max‐stable limit, even for large block sizes.     

Learning,signaling, and convincing: The role of experimentation in the business modeling process

This study examines experimentation in the business modeling process, unpacking three different roles of experimentation: learning, signaling, and convincing. Learning is an inherent role of experimentation, as managers typically experiment to engage with the environment and to obtain knowledge. This study uncovers another set of roles, which have a symbolic nature. These roles show that experimentation is not just a learning process, but also a strategic legitimation process, aimed at enacting the environment. Experimentation serves the purpose of signaling to potential customers and other stakeholders, and of convincing them to embrace the business model. Furthermore, this study shows that experimentation takes two forms—purposeful interactions and experimental projects—and that these forms can support the different roles of experimentation.     

Human Agency in Disaster Planning: A Systems Approach

Current approaches to risk management place insufficient emphasis on the system knowledge available to the assessor, particularly in respect of the dynamic behavior of the system under threat, the role of human agents (HAs), and the knowledge available to those agents. In this article, we address the second of these issues. We are concerned with a class of systems containing HAs playing a variety of roles as significant system elements—as decisionmakers, cognitive agents, or implementers—that is, human activity systems. Within this family of HAS, we focus on safety and mission‐critical systems, referring to this subclass as critical human activity systems (CHASs). Identification of the role and contribution of these human elements to a system is a nontrivial problem whether in an engineering context, or, as is the case here, in a wider social and public context. Frequently, they are treated as standing apart from the system in design or policy terms. Regardless of the process of policy definition followed, analysis of the risk and threats to such a CHAS requires a holistic approach, since the effect of undesirable, uninformed, or erroneous actions on the part of the human elements is both potentially significant to the system output and inextricably bound together with the nonhuman elements of the system. We present a procedure for identifying the potential threats and risks emerging from the roles and activity of those HAs, using the 2014 flooding in southwestern England and the Thames Valley as a contemporary example.     

口感预测型食品物性仿生检测系统

针对目前在使用质构仪进行食品物性检测时，存在的检测数据准确性较差，操作系统功能单一，测试结果与口感 评价间无关联性分析等缺点，提出了基于六自由度并联机器人的口感预测型仿生检测系统。设计了仿生咀嚼检测平台； 利用数理统计原理和最小二乘法建立了物性检测数据与口感评价之间的数学模型，实现了仿生检测系统的口感预测功 能。实验结果表明，利用数学模型获得的口感预测值与人口感评价结果之间的误差小于9%，满足实际应用的需要，验 证了口感预测数学模型构建的正确性及口感预测功能的可行性。     

Macro-Economic Factors in Credit Risk Calculations: Including Time-Varying Covariates in Mixture Cure Models

The prediction of the time of default in a credit risk setting via survival analysis needs to take a high censoring rate into account. This rate is because default does not occur for the majority of debtors. Mixture cure models allow the part of the loan population that is unsusceptible to default to be modeled, distinct from time of default for the susceptible population. In this article, we extend the mixture cure model to include time-varying covariates. We illustrate the method via simulations and by incorporating macro-economic factors as predictors for an actual bank dataset.     

感知风险对在线旅游产品购买意愿的影响分析

旅游电子商务发展迅猛,互联网购物平台的虚拟性和远程性强化了消费者感知风险,它也成为影响旅游电子商务发展的重要因素之一.文章聚焦消费者感知风险,运用SPSS 19.0及AMOS 21.0软件,构建以5个维度的感知风险为自变量、购买意愿为因变量、信任为调节变量的结构方程模型,探讨变量间相关关系.结果发现:感知风险对信任的影响差异较大,仅经济风险感知与信任呈显著负相关;感知风险对购买意愿总效应皆呈负相关;信任作为中介变量存在明显的调节作用,有效地降低了感知风险对购买意愿的负向影响.