[ont.events] U of Toronto Computer Science activities, March 27-31

clarke@csri.toronto.edu (Jim Clarke) (03/11/89)

        ACTIVITIES FOR THE WEEK COMMENCING MARCH 27, 1989
    (SF = Sandford Fleming Building, 10 King's College Road)
         (GB = Galbraith Building, 35 St. George Street)

SUMMARY:

COLLOQUIUM - Tues., March 28, 11 a.m. in Rm SF 1105 -- Donald Norman
     "Cognitive Artifacts, or, Things That Make Us Smart"

NUMERICAL ANALYSIS SEMINAR - Tues., March 28, 11 a.m. in Rm GB 420 -- Paul Muir
     "Error Expressing for Reflected and Averaged Implicit Runge-Kutta Methods"

NUMERICAL ANALYSIS SEMINAR - Fri., Mar. 31, 10 a.m. in Rm SF4102 -- Chris Fraley
     "Single-Phase vs. Multi-Phase Projective Methods for Linear Programming"

--------------------

COLLOQUIUM - Tuesday,  March 28,  11 a.m. in  Room  SF 1105

                          Donald Norman
               University of California, San Diego

      "Cognitive Artifacts, or, Things That Make Us Smart"

The power of the unaided human mind is highly overrated.  Ar-
tifacts play a critical role in human performance, whether it be
as an aid to memory, spatial reasoning, attentional focus, or
communication.  I give examples of the role that even simple ar-
tifacts can play -- for spatial communication, reminders, pre-
computation, task restructuring -- and present the beginnings of
a theoretical analysis of the interaction between internal and
external knowledge and structure.

NUMERICAL ANALYSIS SEMINAR - Tuesday, March 28,  11:10 a.m. in
Room  GB 420

                            Paul Muir
                     Saint Mary's University

                      "Error Expressing for
      Reflected and Averaged Implicit Runge-Kutta Methods"

In addition to their usefulness in the numerical solution of ini-
tial value ODE's, the implicit Runge-Kutta (IRK) methods are also
important for the solution of two-point boundary value problems.
Recently, several classes of modified IRK methods which improve
significantly on the efficiency of the standard IRK methods in
this application have been presented. One such class is the Aver-
aged IRK methods; a member of the class is obtained by applying
an averaging operation to a non-symmetric IRK method and its re-
flection. In this talk we investigate the forms of the error ex-
pressions for reflected and averaged IRK methods. Our first
result relates the expression for the local error of the reflect-
ed method to that of the original method. Our main result relates
the error expression of an averaged method to that of the method
upon which it is based. We apply these results to show that for
each member of the class of the averaged methods, there exists an
embedded lower order method suitable for use in error estimation,
in a formula-pair fashion.

NUMERICAL ANALYSIS SEMINAR - Friday, March 31,  10 a.m.  in Room
SF 4102

                          Chris Fraley
                      University of Geneva

        "Single-Phase vs. Multi-Phase Projective Methods
                     for Linear Programming"

This talk concerns a specific class of projective methods for
linear programming.  The methods in this class differ from most
projective methods in that they never project onto the unit sim-
plex, and they need not be started at a feasible point for the
linear program.  More than one algorithm can be devised within
this framework.  For example, it is possible to treat feasibility
and optimality in entirely separate phases, or to approach feasi-
bility and optimality simultaneously.  We discuss several dif-
ferent options, and give comparative performance results.  We
will also address some general computational issues related to
projective methods for linear programming.
-- 
Jim Clarke -- Dept. of Computer Science, Univ. of Toronto, Canada M5S 1A4
              (416) 978-4058
clarke@csri.toronto.edu     or    clarke@csri.utoronto.ca
   or ...!{uunet, pyramid, watmath, ubc-cs}!utai!utcsri!clarke