Abstract: We observe periodic phenomena everyday in our lives. The daily temperature of Delhi or the number of tourists visiting the famous Taj Mahal or the ECG data of a normal human being, clearly follow periodic nature. Sometimes, the observations may not be exactly periodic due to different reasons, but they may be nearly periodic. The received data is usually disturbed by various factors. Due to random nature of the data, statistical techniques play important roles in analyzing the data. Statistics is also used in the formulation of appropriate models to describe the behavior of the system, development of an appropriate technique for estimation of model parameters, and the assessment of model performances. In this talk we will discuss different techniques which have developed for the last twenty five years for analyzing periodic data, other than the standard Fourier analysis.
Venue
LH 101
Speaker
Professor Debasis Kundu
Affiliation
Department of Mathematics & Statistics, Indian Institute of Technology, Kanpur
Title
Analyzing Periodic Data: Statistical Perspectives