COURSE INFORMATION

550.695 ADVANCED PARAMETERIZATION IN SCIENCE AND ENGINEERING

FALL 2011

Instructor: Gregory Eyink
Department: Applied Mathematics & Statistics
Office: Whitehead 202-D
Office hours: Mon-Wed, 12:30-1:20pm
Email: eyink@jhu.edu
Phone: (410) 516-7201


Classroom:    Maryland 104

Course Web Page:    http://www.ams.jhu.edu/~eyink/Advance-Param

Text:   The notes of the lecturer and cited journal articles will constitute the primary source for the course.

Overview:   This course is a continuation of 560.702, ``Modeling Complex Systems Colloquium.''
(See also 560.700, ``Applications of Science-Based Coupling of Models'', linked on the main page.)
Although not strictly a prerequisite, the IGERT Colloquium 560.702 shares the main subject and goals
of the present course. The basic topic is the coupling of multi-scale/multi-physics models in the context of
a broad range of science and engineering applications, emphasizing calculable approximation schemes.
The course is fundamentally interdisciplinary. Two of its key objectives are to help students to develop an
appreciation for the many disciplines in which science-based coupling of models is important and also to
help students communicate with researchers in other areas, in order to share and cross-fertilize ideas
and methods.

The principal difference of this course from 560.702 is methodological. In the Colloquium, Hopkins IGERT
faculty from diverse fields present a series of broad themes in modeling complex systems together with
sample applications, discussing the problems, methods, and prospects in those several areas. In our
discussions, however, the emphasis will be on unity as well as diversity. An attempt will be made to
identify common mathematical structures and methods that cut across the several fields of research.
A deeper understanding of the shared mathematics and tools should permit more effective communication
between researchers and better ability to adapt and exploit new approaches.

Topics Covered:

I.    Dynamics.

II.    Coarse-Grained Dynamics.

III.    Ensembles & Probability.

IV.    Moment Closure.

V.    Statistical Estimation & Uncertainty Quantification.

VI.    Hybrid/Multiscale Algorithms.

Grading:

Your final grade in this class will be determined from homework and from participation in scheduled
classroom conference sessions.

There will be periodic homework assignments, with written solutions due in class at the beginning of
conference sessions. Most of the homework problems will fill in details of lecture examples, with the
aim of deepening students' understanding. Students are encouraged to discuss homework, but all solutions
must be written up and submitted individually. The instructor's solutions to assigned exercises
will be presented in the conference sessions and on-line.

In addition to formal homework assignments, students in the course have an important additional on-going
assignment. As each topic is discussed in the course, students are asked to think about how the material
is relevant (or irrelevant!) to their own research interests. These thoughts should be recorded in writing
at the Google Group "JHUmultiscale" discussion pages:

http://groups.google.com/group/jhumultiscale

before the in-class conference session on that topic. Discussion of the lecture material in the
conference sessions will be an integral part of the course.