-
2 DAYS WORKSHOP ON THE INTRODUCTION OF
-
ROBUST STATISTICS AND ITS APPLICATIONS
In classical set setup, the assumptions that are common to almost all statistical tests are that the observations are random, independent and identically distributed, come from a normal distribution and equally reliable (there is no outlier in a data). Outliers are observations which are markedly different or far from the majority of observations. In statistical data analysis, there is only one type of outlier, but in a regression problem, extra care should be taken because in this situation, there are several versions of outliers exist such as residual outliers, vertical outliers and high leverage points. The classical methods heavily depend on assumptions and the most important assumption is that data are normally distributed. However, in practice those assumptions are difficult to be met. Violations of at least one of the assumptions may produce sub-optimal or even invalid inferential statements and inaccurate predictions. The immediate consequence of outlier is that it may cause apparent non-normality and the entire classical methods might breakdown. Even one single outlier can have arbitrarily large effect on the estimates.
Since outliers give bad consequences, the need for robust methods become essential to avoid misleading conclusion. Robust statistics are those statistics that do not breakdown easily. It is less affected by outliers by keeping its effect small.
In this workshop participants will be introduced to robust statistical methods as an alternative to the classical methods which are easily affected by outliers. Participants will have the opportunity to learn R programming language and will be guided on theinterpretation of the outputs obtained.
After attending this course, the participants should be able to use some identification methods for the detection of outliers and high leverage points and also able to employ robust methods to analyse a data set. Participants will also have the opportunity to obtain statistical consulting service/statistical advice on the last day of the workshop.
Who Should Attend ?
Academicians, Researchers, Graduate Students and Researchers from various industries both public and private.
