Note: All of the following seminars are located at MEB 103 and take place from 12:30 pm until 1:50 pm unless otherwise stated.
Thursday, January 6th
Speaker : Anne Greenbaum and others
Title : ACMS Organizational Meeting
Thursday, January 13th
Speaker :
William Stein, UW Math
Title : An Introduction to the Sage Open Source Mathematical Software Project
Abstract : I started the Sage mathematics software project
(http://www.sagemath.org) in 2005, and since then many students have
used Sage and also contributed code and documentation. Come to this
talk and find out more about how Sage is useful to you, and how you
might contribute back.
Click Here for the Seminar webpage.
Thursday, January 20th
Speaker : Chris Swierczewski,UW Applied Math Graduate Studetn
Title : Polynomial Approximations to Functions
Abstract : In calculus we learned about approximating functions using their
Taylor expansions. Although terribly useful in analysis they're a terribly useless in
numerical computation. For example, in order to approximate a function on an entire
interval to high accuracy we often need to take so many terms in the Taylor series that
we lose any computational speed benefits. Lagrange interpolation is another approach to
approximating functions using polynomials but the naive approach has its issues as well.
We will introduce a new technique developed by Nick Trefethen that approximates functions
using "Chebyshev polynomials" to high accuracy and with lower computational complexity.
These approximations, called "Chebfuns", have various analytic properties that allow us
to quickly and accurately compute numerical integrals and derivatives as well. Many pretty
pictures to share, too!
Click Here for talk slides
Thursday, January 27th
Speaker : Andrew F. Siegel
Title : Arbitrage-Free Linear Price Function Models for the Term Structure of
Interest Rates
Abstract:These arbitrage-free term structure models represent the time evolution of
bond prices - not yields or forward rates - as a finite linear combination
of fixed functions of the time to maturity. These models are less complex
than affine models, the differential equation admits a general closed-form
solution, and specification of conditional stochastic volatility (and
correlation) for the state variables is very general because the drift is
not affected by the diffusion. A specific model is derived within this
family that closely mimics the widely-used Nelson-Siegel yield curves,
demonstrating that practical, parsimonious models are available within this
larger class.
Thursday, February 3rd
Speaker : Fritz Scholz, UW STAT, Retired from The Boeing Company
Title : Probability and Statistics Applications in Aviation and Space
Abstract: This talk will skim over many different application examples
involving statistics and probability as I encountered them during my 28
years with The Boeing Company. Typically I will hint at the issues
involved and occasionally go into more depth. The main purpose is to
convey the breadth of application range.
Click Here for talk slides
A few other talks (mentioned in seminar) and material can be found at the speaker's website.
Thursday, February 10th
Speaker : Emo Todorov
Title : Optimal Control of Movement
Abstract:
Thursday, February 17th
Speaker : Krzysztof Burdzy, UW MATH
Title : Financial mathematics and stochastic calculus
Abstract:I will discuss some mathematical ideas
related to financial mathematics and stochastic
processes. The presentation will be totally
informal and accessible to all undergraduate
students. The talk will effectively be an advertisement
for the course Math 492 "Stochastic Calculus for
Option Pricing".
Thursday, February 24th
Speaker : Robin Evans
Title : Probabilistic Causal Models
Abstract:
Correlation does not imply causation - so how do we know what causes what? Our scientific instincts are are not satisfied by finding mere correlations or 'associations' between different events or variables; we want to know why things are associated. So how, if at all, can statistics be used to infer causal connections? We give a brief overview of some of the techniques used, and how our research contributes to solving these problems.
This is joint work with Thomas Richardson.
Thursday, March 3rd
Speaker : John Palmieri, UW MATH
Title : The Brouwer fixed point theorem
Abstract:The Brouwer fixed point theorem is an important topological result,
with connections to various other fields of mathematics. I will
discuss the theorem and some of its applications, and I will sketch a
proof using tools from combinatorics.
Talk slides
Thursday, March 10rd
Speaker : Brad Bell, UW Applied Physics Lab
Title : An Introduction by Example to Algorithmic Differentiation
Abstract:Algorithmic Differentiation (often referred to as Automatic Differentiation or just AD) uses the software representation of a function to obtain an efficient method for calculating its derivatives. These derivatives can be of arbitrary order and are analytic in nature (do not have any truncation error). A forward mode sweep computes the partial derivative of all the dependent variables with respect to one independent variable (or independent variable direction). A reverse mode sweep computes the derivative of one dependent variable (or one dependent variable direction) with respect to all the independent variables. The number of floating point operations for either a forward or reverse mode sweep is a small multiple of the number required to evaluate the original function. Thus, using reverse mode, you can evaluate the derivative of a scalar valued function with respect to thousands of variables in a small multiple of the work to evaluate the original function.
AD automatically takes advantage of the speed of your algorithmic representation of a function. For example, if you calculate a determinant using LU factorization, AD will use the LU representation for the derivative of the determinant (which is faster than using the definition of the determinant).
Talk on the web