This is a reminder about today's seminar at UMIST.
The details are given below. All those interested are welcome to attend.
Wednesday 24th November, 2.00pm - 3.00pm
Venue: UMIST, MSS/O10. Note change of venue.
Speaker: Dr Jingsong Yuan, UMIST.
Title: Testing Linearity for Stationary Time Series Using the Sample
Interquartile Range
Abstract:
We present theoretical results on the size and power of Hinich's linearity
test and propose a modification. The modified test also uses the sample
interquartile range, but it is based on normal distributions rather than the
noncentral chi-squared. It can be interpreted as testing the equality of
location parameters. The null distribution is known and the critical value
particularly simple to calculate. The limitation of the sample interquartile
range as a measure of deviation from the null hypothesis is discussed.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|