CFE 2019: Start Registration
View Submission - CMStatistics
B0745
Title: Pseudo-binomial Approximation to $(k1, k2)$-runs Authors:  Amit Kumar - Indian Institute of Technology Bombay (India) [presenting]
Neelesh Shankar Upadhye - Indian Institute of Technology Madras (India)
Abstract: It is known that the distribution of $(k1, k2)$-runs, arising from non-identical Bernoulli trials, is intractable. Hence, it is important to approximate $(k1, k2)$-runs by a suitable distribution, with reasonable accuracy. We discuss pseudo-binomial approximation to $(k1, k2)$-runs using Stein's method. The approximation results are of optimal order and improve upon the existing results in the literature.