d e f es 
AICOS Technologies at PCH Meetings 2020 in Lyon
News PCH Lyon March, 2020 
más Linie
Datos de primavera cursos en línea !
La agenda de cursos para la primera mitad del año... 
más Linie
Here you can read the latest news about products,... 
más Linie

Analysis of Stability Data

Thu/Fri, 4-5/6/2020

Disculpe las molestias, la página solicitada no está disponible en español.

The course covers the treatment of censored data and statistical basic concepts (survival function, hazard rate) as well as the most common parametric and non-parametric lifetime and regression models (exponential and Weibull distribution, Kaplan-Meier estimator, Cox model).

What will you learn ?
Are you in charge of making sure that your products work without failure or keep constant properties during a fixed period of time? The most common methods for reliability and lifetime data analysis are covered in this course and will help you to estimate reliably the lifetime or the stability of your product. The emphasis is on the use of graphical methods and the choice and fit of convenient lifetime distributions. The methods will be explained using practical examples and exercises that stem from the chemical, pharmaceutical and machine industries as well as software demonstrations on PC. The implementation of the ICH-Guidelines will be shown in detail.

Who should attend ?
  • Quality managers, chemists, pharmacists and engineers who are responsible for the lifetime quality or the stability of products
  • People who have to analyse censored and uncensored data
  • Elementary statistical knowledge is assumed (as given in the course "Visualization of Lab Data" )

Which topics are coveres ?
 Basic Concepts
 Data Situation
Basic concepts of lifetime data analysis
 Statistical models for stability data (main issue)
 Linear regression:
  • Parameter estimation
  • Confidence domains
Covariance analysis
Data "Pooling"
Treatment of several responses
Implementation of ICH-Guidelines
 Statistical models for lifetime data
 Survival plots
Hazard plots
Parametric models:
  • Exponential distribution
  • Weibull distribution
Non-parametric models:
Kaplan-Meier estimator
Accelerated-Failure-Time model
Cox model

Any questions ?
powered by Swisscom AG