STAT - 570: Stochastic Processes.  Spring 2004.

 

 

SYLLABUS.                     HOMEWORK                   HANDOUTS     

 

NEWS

                                     

Back to the teaching page._______________

 

Instructor:  Professor Vladimir Rotar; Office GMCS -514, phone: 594 7244; e-mail: rotar@sciences.sdsu.edu or vrotar@euclid.ucsd.edu.

 

Classes:           MW, 14:00-15:15, GMCS-314.

Office hours:   MW, 12:45-13:45 and 16:00-17:00; F, 12-13; or by appointment.

           

Text: Howard Taylor and Samuel Karlin  (1998).  An Introduction to Stochastic Modeling, 3rd Edition  Academic Press.

           

Other References:

  1. Sheldon Ross, (1997), Probability Models, 6th edition, Academic Press.
  2. Vladimir Rotar, (1998), Probability Theory, World Scientific. 

           

Examinations. There will be many quizzes, and a final exam. Homework will be assigned almost each class and will be collected and graded, as a rule, each week.)

 

      This course may be viewed as a continuation of the introductory probability course. The main step we take here consists in considering random phenomena in their dynamics. Not just the number of people who will live in San Diego in the next year (which is a random variable) but the process of the San Diego population growth during a certain period (and a tendency of this growth); not just the tomorrow stock prices but their evolution in the long run; not just the number of customers a company will deal with tomorrow but the flow of customers during a particular season; etc., etc. 

      What do we study from a mathematical point of view? The graph of, say, the air temperature against time during the whole day is a curve; if this day is tomorrow, it is a random curve, or a random process. So, while the object of study in the first course on Probability is random variables, the object of study in Random Processes course is random functions. 

 

           Three main questions arise. 

·        How or in what terms should we describe random processes (or random functions, which is the same), what new notions should we introduce? 

·        What kinds of processes are most important, how to classify them?

·        What qualitatively new phenomena can we watch when we switch from the static to the dynamic framework? What qualitatively new and interesting facts can we discover? 

 

A good example of such facts is ergodicity. For example, the tomorrow temperature strongly depends on the today temperature, the dependence of the today temperature and that in a week is much weaker, while the temperature, say, in three months practically does not depend on what temperature we have today. Thus the system “is forgetting the past”, the influence of the starting state is vanishing in time. When it is the case, under what conditions we can watch such phenomena is one of the main questions of the stochastic processes theory.

 

The prerequisite is ordinary Calculus and Linear Algebra (not complicated, but the  student should know how to differentiate and integrate, how to multiply matrices, and what their eigenvalues are), and an introductory course on Probability Theory (say, Stat-550 or Stat-551a are  enough).