Density forecasting for failure probability in manufacturing
by Ilmari Juutilainen; Satu Tamminen; Juha Röning
European J. of Industrial Engineering (EJIE), Vol. 9, No. 4, 2015

Abstract: Density forecasting is a subfield of multivariate regression aimed at accurate prediction of full conditional distributions. This article presents methods for improving product quality by deploying density forecast-based failure probability predictors that predict the risk of failure to meet the requirements of qualification tests and specification limits. Algorithms that efficiently deploy failure probability predictors in target optimisation problems and in process monitoring, planning and control operations are provided. In one of the three case applications, density forecast methods decreased production costs more efficiently than the reference method, i.e., point prediction for mean. In two case applications, density forecast methods did not provide additional value. To promote exploitation of density forecasting, the article presents ideas and prototype implementations for integrating density forecast-based failure probability predictors into software applications employed to improve the production efficiency of manufacturing processes. [Received 17 March 2013; Revised 22 October 2013; Revised 8 March 2014; Accepted 6 April 2014]

Online publication date: Thu, 02-Jul-2015

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the European J. of Industrial Engineering (EJIE):
Login with your Inderscience username and password:

    Username:        Password:         

Forgotten your password?


Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com