Profile Monitoring with Binary Data and Random Predictors

• Autores: Yanfen Shang, Fugee Tsung, Changliang Zou
• Localización: Journal of quality technology: A quarterly journal of methods applications and related topics, ISSN 0022-4065, Vol. 43, Nº. 3, 2011, págs. 196-208
• Idioma: inglés
• Enlaces
  • Texto completo
• Resumen

  • Profile monitoring checks the stability over time of relationships between explanatory and response variables. A control chart is proposed for profiles with random explanatory variables. This proposal integrates an exponentially weighted moving-average scheme and a likelihood ratio test. The proposed scheme is reasonably noncomplex, and simulations show that it performs better than standard methods. A real example from the electronics industry also shows the utility of this approach.