This paper describes a new and simple design procedure for life testing based on the Kullback-Leibler (hereafter K-L) information. Information theory is a special branch of probability theory. The K-L information is utilized as a criterion for measurement of the distance between the observed and the model probability distributions. In this paper we make the assumption that the length of life has a Weibull distribution. In the reliability literatures, the Weibull distribution is a popular parametic family of life-time distributyion. In the fixed time testing, it is attempted to determine the optimum censored time from the economic viewpoint.
抄録全体を表示