![]() ![]() ![]() Zero inflated models are used to model count data for which proportion of zero counts is greater than expected. However many data sets in general are not well fitted by a Poisson model, because they consists of more zero counts than are compatible with the Poisson model. In this juncture there is a considerable procedure to deal with this problem using Poisson distribution which is often used as a standard probability model for dealing with this kind of single sampling plan. The trade-off is that the Dodge-Romig tables provide a sampling plan for a range at lot sizes and process averages. Determining these plans from basic principles would take much more time. If the parameters of the plan under consideration (eg., lot size and process average) are within certain range, the Dodge – Romig tables allow us to determine feasible plans very readily. For both sets of plans, the objective is to minimize the average total inspection. The other based on meeting a certain value of the average outgoing limit. One is based on satisfying a given limiting quality level (LQL) based on a consumers risk β, the target value of which is 0.10 (Dodge and Romig, 1959). Dodge – Romig plans can be single or double plans. Acceptance sampling systems are advocated when small sample size are necessary or desirable towards costlier testing for product quality. Abstract: Acceptance sampling procedures are the practical tools for quality assurance applications involving product control.
0 Comments
Leave a Reply. |