Page 327 - Computational Statistics Handbook with MATLAB

316 Computational Statistics Handbook with MATLAB

8.17. Repeat Example 8.12. Plot the curves from the estimated models.
What is the ISE between the two estimates? Use the iterative EM
algorithm on both models to refine the estimates. What is the ISE after
you do this? What can you say about the two different models? Are
your conclusions different if you use the IAE?
8.18. Write a MATLAB function that will generate random variables
(univariate or multivariate) from a finite mixture of normals.
8.19. Using the method for generating random variables from a finite
mixture that was discussed in this chapter, develop and implement
an algorithm for generating random variables based on a kernel den-
sity estimate.
8.20. Write a function that will estimate the MISE between two functions.
Convert it to also estimate the MIAE between two functions.
8.21. Apply some of the univariate density estimation techniques from
this chapter to the forearm data.
8.22. The elderly data set contains the height measurements (in centi-
meters) of 351 elderly females [Hand, et al., 1994]. Use some of the
univariate density estimation techniques from this chapter to explore
the data. Is there evidence of bumps and modes?
8.23. Apply the multivariate techniques of this chapter to the nfl data
[Csorgo and Welsh, 1989; Hand, et al., 1994]. These data contain bivari-
ate measurements of the game time to the first points scored by
), and the game time to
kicking the ball between the end posts ( X 1
).
the first points scored by moving the ball into the end zone ( X 2
The times are in minutes and seconds. Plot your results.

© 2002 by Chapman & Hall/CRC