International Journal of Reliability, Quality and Safety Engineering

This paper presents the multi-based experts Failure Effects Analysis (FEA). The experts' opinions differ substantially because the experts do not often agree on the level of the failure factors (failure probability, non-detection probability, severity of effect, and expected cost) and the functions/subsystems attributes (e.g., importance). Therefore, conflict always occurs in Group-based Failure Effects Analysis (GFEA). The approach uses fuzzy Risk Priority Category (RPC) and group decision-making techniques to study both the failure effects on the functions/subsystems and the failure risk category with uncertain information. In addition, the approach uses the compensated operators to allow the tradeoffs either among failure factors or among functions/subsystems attributes. A solved example is presented to demonstrate the Group-based Failure Effects Analysis (GFEA) application.

A support vector machine (SVM) modeling approach for software reliability prediction is proposed. Based on the structural risk minimization principle, the learning scheme of SVM is focused on minimizing an upper bound of the generalization error that eventually results in better generalization performance. The SVM learning scheme is applied to the failure time data, forcing the network to learn and recognize the inherent internal temporal property of software failure sequence. Further, the SVM learning process is iteratively and dynamically updated after every occurrence of new failure time data in order to capture the most current feature hidden inside the software failure behavior. The performance of our proposed approach has been tested using four real-time control and flight dynamic application data sets and compared with feed-forward neural network and recurrent neural network modeling approaches. Experimental results show that our proposed approach adapts well across different software projects, and has a better next-step prediction performance.

In this paper, the author discusses estimation procedures for K independent, homogeneous geometric processes (GPs), parallel to those for K independent, homogeneous arithmetic processes (APs) derived by Leung.³ In addition, he corrects two minor mistakes appearing in Leung,³ on pp. 231–232. In the course of developing the fourth and fifth estimators for , the mean of , the author was inspired to derive two additional corresponding estimators for , the mean of . For the reader's easy reference, the results of both the APs derived in Leung³ and GPs derived in this paper are summarized in a table. Finally, the author draws some conclusions of the applicability of an AP and/or a GP based on partial findings of two real case studies.

Deceleration in mortality rates for old ages is explained via the concept of population heterogeneity. Two simple probabilistic models of biological aging are considered. The first one assumes that some random resource is acquired by an organism at birth. Death occurs when the accumulated wear exceeds the initial random resource. In the second model death occurs as the consequence of a harmful event. A non-homogeneous Poisson and doubly stochastic Poisson processes of harmful events are considered. These models describe possible causes of population heterogeneity.

This paper provides an alternative valuation approach to reliability design based a concept of Value at Risk (VaR) used in finance to measure risk exposure. An accounting of the direct and indirect costs combined with the costs needed to meet contingent claims in case of default is used to redefine the traditional reliability design problem, resulting is a nonlinear optimization problem which we can solve by the usual methods. Applications are used to demonstrate the approach used.

A problem of individual prediction of the state of technical systems is considered, where an approach for solving the problem is based on construction of minimax robust algorithms applicable in case of incomplete and insufficient measurements. Suggested algorithms are tested with real data samples.