Discrimination between accelerated life models via Approximate Bayesian Computation

Accelerated life testing (ALT) is widely used in high-reliability product estimation to get relevant information about an item's performance and its failure mechanisms. To analyse the observed ALT data, reliability practitioners need to select a suitable accelerated life model based on the nature of the stress and the physics involved. A statistical model consists of (i) a lifetime distribution that represents the scatter in product life and (ii) a relationship between life and stress. In practice, several accelerated life models could be used for the same failure mode and the choice of the best model is far from trivial. For this reason, an efficient selection procedure to discriminate between a set of competing accelerated life models is of great importance for practitioners. In this paper, accelerated life model selection is approached by using the Approximate Bayesian Computation (ABC) method and a likelihood-based approach for comparison purposes. To demonstrate the efficiency of the ABC method in calibrating and selecting accelerated life model, an extensive Monte Carlo simulation study is carried out using different distances to measure the discrepancy between the empirical and simulated times of failure data. Then, the ABC algorithm is applied to real accelerated fatigue life data in order to select the most likely model among five plausible models. It has been demonstrated that the ABC method outperforms the likelihood-based approach in terms of reliability predictions mainly at lower percentiles particularly useful in reliability engineering and risk assessment applications. Moreover, it has shown that ABC could mitigate the effects of model misspecification through an appropriate choice of the distance function.     

Towards an ICT-based psychology: E-psychology

Cognitive science is the scientific domain which studies, analyzes, simulates and infers for various aspects, functions and procedures of human mentality such as, thinking, logic, language, knowledge, memory, learning, perception and the ability to solve problems. E-psychology is in close relation with the cognitive science domain, but expands beyond it, as e-psychology is the efficient convergence of psychology and Information and Communication Technologies (ICTs). E-psychology offers a number of services such as supporting, diagnosis, assessment, therapy, counseling, intervention and tests through an effective exploitation of ICTs. This article presents a user-friendly, flexible and adaptive electronic platform, which supports both synchronous and asynchronous e-psychology activities through the use of informative and communicative tools and services, which can be adapted to support various methods of e-psychology activities. It is important to underline that e-psychology is not an alternative psychology field, but a resource to enhance the conventional psychology process.     

Formal model for assigning human resources to teams in software projects

Context

Human resources play a critical role in software project success. However, people are still the least formalized factor in today’s process models. Generally, people are assigned to roles and project teams are formed on the basis of project leaders’ experience of people, constraints (e.g. availability) and skill requirements. Yet this process has to take multiple factors into account. Few works in the literature model this process. Most of these are informal proposals focusing on the individual assignment of people to project tasks and do not consider other aspects like team formation as a whole.

Objective

In this paper we formulate a formal model for assigning human resources to software project teams. Additionally, we describe the key results of the knowledge management process enacted to output the elements of the model.

Method

The model elements were identified using the Delphi expert consultation method and applying psychological tests. The proposed model was implemented in a software tool and validated on two software development organization assignment scenarios.

Results

We built a formal model for the process of assigning human resources to software project teams. This model takes into account as many factors as possible and aids the assignment of individuals to project roles, as well as the formation of the team as a whole.We found that the rules that were identified to form software development project teams are useful. From the tests we found that model implementation was feasible (all the executions of the implemented problem-solving algorithms output feasible solutions in response times that can be considered as acceptable).

Conclusion

Using the Delphi method we were able to propose software project roles and competences. Psychological tests and data mining tools identified useful rules for forming software project teams. These were used to build a formal model. This model was built into a tool that returns role assignments in acceptable response times. This decision support tool helps managers assign people to roles and to form teams. Using the tool, project leaders can flexibly evaluate different team make-ups, taking into account several factors, as well as different constraints and objectives.     

A bootstrap goodness of fit test for the generalized Pareto distribution

This paper proposes a bootstrap goodness of fit test for the Generalized Pareto distribution (GPd) with shape parameter γ. The proposed test is an intersection–union test which tests separately the cases of γ≥0 and γ<0 and rejects if both cases are rejected. If the test does not reject, then it is known whether the shape parameter γ is either positive or negative. A Monte Carlo simulation experiment was conducted to assess the power of performance of the intersection–union test. The GPd hypothesis was tested on a data set containing Mexico City’s ozone levels. ¹     

A study on the use of statistical tests for experimentation with neural networks: Analysis of parametric test conditions and non-parametric tests

In this paper, we focus on the experimental analysis on the performance in artificial neural networks with the use of statistical tests on the classification task. Particularly, we have studied whether the sample of results from multiple trials obtained by conventional artificial neural networks and support vector machines checks the necessary conditions for being analyzed through parametrical tests. The study is conducted by considering three possibilities on classification experiments: random variation in the selection of test data, the selection of training data and internal randomness in the learning algorithm.The results obtained state that the fulfillment of these conditions are problem-dependent and indefinite, which justifies the need of using non-parametric statistics in the experimental analysis.     

Robust Estimation of GARMA Model Parameters with an Application to Cointegration among Interest Rates of Industrialized Countries

We review the literature on long memory ARFIMA and GARMA models andintroduce a new efficient estimator for GARMA models, which we show to berobust. Next we conduct a Monte Carlo study to demonstate the power of theDickie–Fuller test when the data are generated from a stationary GARMAprocess. We conclude with a brief discussion of cointegration in the contextof GARMA models with an application to international interest rates.     

A review of: “Introduction A La Psychophysiologie Du Travail (Introduction to the Psychophysiology of Work)”, by ROBERT FLORU and JEAN-CLAUDE CNOCKAERT,Presses Universitaires De Nancy (1991). pp. vi + 233, FF 120, ISBN 2-86480-550-2.

Workers in physically demanding occupations (PDOs) are frequently subjected to physical selection tests. To avoid legal ramifications, workplaces must be able to show that any personnel selection procedures reflect the inherent requirements of the job. A job task analysis (JTA) is fundamental in determining the work tasks required for employees. To date, there are no published instructions guiding PDO researchers on how to conduct job task analyses. Job task analysis research for non-PDOs offers some insight into the expected reliability and validity of data obtained on the most prevalent task domains in job analysis (importance, frequency, time spent and difficulty). This review critiques such research, and the existing published material on JTA of PDOs, and provides recommendations for future research and practice.

Practitioner Summary: There are no published guidelines for physically demanding occupation (PDO) researchers conducting job task analysis (JTA). Given the legal consequences of improperly conducted JTA, scientifically valid instructions for JTA practitioners are required. This review critiques existing research which analyses reliabilities of JTA data, and provides guidelines for PDO researchers conducting JTA.     

服装舒适性主观评价及主客观指标间的相关性研究

针对评估主观测试方法在服装舒适性上的应用,本文运用数理统计的方法,对涤纶、棉、涤/棉混纺面料服装在不同的气候条件下对其舒适性进行主观测试评价,并对其主要物理指标进行测试,同时运用回归分析对主观舒适性指标和客观测试指标进行了相关性分析,实验结果表明,主客观测试能够辨别出服装舒适性能的总体差异,并且对于服装舒适性趋势预测的呈现有所帮助。主客观测试结果具有一定的相贯性,而与服装舒适性主观感觉相关度最高的测试指标为水蒸汽传递速率、透气率、织物模量以及悬垂指数。主观测试方法可以很好的适用于热、冷环境下着装人体热湿舒适性的研究。