 |
|
Event(s) on December 2005
- Tuesday, 6th December, 2005
| Title: |
Least absolute deviation estimation for fractionally integrated autoregressive moving average time series models wit conditional heteroscedasticity |
| Speaker: |
Prof. Wai Keung LI, Department of Statistics and Actuarial Science, The University of Hong Kong |
| Time/Place: |
11:30 - 12:30
FSC1217
|
| | |
| Abstract: |
In order to model time series exhibiting the features of long
memory, conditional heteroscedasticity and heavy tails, a least
absolute deviation approach is considered to estimate fractionally
autoregressive integrated moving average models with conditional
heteroscedasticity. The time series generated by this model is
short memory or long memory, stationary or nonstationary, depending
on whether the fractional differencing parameter d (-1/2,0)
or (0, ∞ ), (-1/2,1/2) or (1/2, ∞) respectively.
Using a unified approach, the asymptotic properties of the least
absolute deviation estimation are established. This article also
derives the large sample distribution of residual autocorrelations
and absolute residual autocorrelations and these results lead
to two useful diagnostic tools for checking the adequacy of the
fitted models. Some Monte Carlo experiments were conducted to
examine the performance of the theoretical results in finite
sample cases. As an illustration, the process of modeling the
absolute return of the daily closing Dow Jones Industrial Average
Index (1995-2004) is also reported.
|
- Friday, 16th December, 2005
| Title: |
On Surrogate Dimension Reduction for Measurement Error Regression: An Invariance Law |
| Speaker: |
Professor YIN, Xiangrong, Department of Statistics, University of Georgia, USA |
| Time/Place: |
14:30 - 15:30
FSC1217
|
| | |
| Abstract: |
It has been recently discovered that several well known dimension
reduction methods, such as OLS, SIR, and pHd, can be performed
on the surrogate regression problem to produce correct estimate
for the original regression problem involving the unobserved
true predictor. In this paper we will establish a general invariance
law between the surrogate and the original dimension reduction
spaces which implies that, at least at the population level,
the two dimension reduction problems are in fact equivalent.
This allows us to use the recently developed dimension reduction
techniques to tackle the difficult situations in which the classical
methods are inaccurate. We apply several dimension reduction
methods to real and simulated data sets involving measurement
error to compare their performances.
|
- Friday, 16th December, 2005
| Title: |
The State of Mathematics Research and Doing Postgradute Research in Germany |
| Speaker: |
Profs.R.Rannacher/C.Carstensen, University of Heidelberg/Humboldt University |
| Time/Place: |
16:00 - 17:30
ACC 109
|
| | |
| Abstract: |
Sponsored by The Croucher Foundation, DAAD and Hong Kong Baptist
University.
|
- Monday, 19th December, 2005
| Title: |
The role of nonlinearities in the biophysical properties and biological function of key subcellular components |
| Speaker: |
Prof. Jack Tuszynski, Department of Physics, The University of Alberta |
| Time/Place: |
11:00 - 12:00
FSC1217
|
| | |
| Abstract: |
This paper discusses the role of nonlinearities in the physical
description of several key biomolecules that participate in a
number of crucial subcellular processes, namely actin, microtubules
and ions crowding around these filaments. We show that the assembly
kinetics of actin is a nonlinear process that requires not only
a mechanism of saturation but also annealing and fragmentation
that are governed by coupled nonlinear equations involving monomer
concentration and filament number as the key dynamical variables.
The observed dendritic growth of actin networks in cell motility
phenomena is subsequently described by the coupling of actin
filaments to the protein called Arp 2/3. We then investigate
the role of nonlinear dynamics in the formation of microtubules.
First of all, space-flight laboratory experiments have shown
that the in vitro and in vivo self-organization of microtubules
is sensitive to gravitational conditions. We propose a model
of self-organization of microtubules in a gravitational field.
The model is based on the dominant chemical kinetics. The pattern
formation of microtubule concentration is obtained: 1) in terms
of a moving kink in the limit when the disassembly rate is negligible,
and 2) for the case of no free tubulin and only assembled microtubules
present. The results of our simulations are in good quantitative
agreement with experimental data. Next, we present a recently
proposed model of molecular and bulk elastic properties of microtubules
that include macroscopic estimates of the anisotropic elastic
moduli of microtubules, accounting for the molecular forces between
tubulin dimers: for a longitudinal compression, for a lateral
force and for a shearing force. At the level of large bending
motions of microtubule filaments, a continuous medium model is
proposed describing a microtubule as an elastic rod. Keeping
the dominant nonlinear terms in the bending dynamics equation,
we found that when the microtubule is subjected to bending forces,
the deviation angle satisfies a Sine-Gordon equation. Particular
analytical solutions of this equation are found which describe
kink and anti-kink bending modes that may propagate at a range
of velocities along the length of the microtubule. Kinetic energies
and characteristic damping times of these modes are calculated
for different propagation velocities and compared with thermal
and ATP hydrolysis energies. Finally, we discuss how coupled
differential equations describing the interactions between ions
in solution and the filament they surround can lead to solitonic
signal transmission.
This work was supported by grants from MITACS and NSERC awarded
to J.A.T. and S.P. acknowledges support of the Bhatia post-doctoral
fellowship.
|
- Wednesday, 21st December, 2005
| Title: |
Combining Data from different Sources: for enhancement of data value and removal of bias using geo-demographic clusters |
| Speaker: |
Jae Chang Lee, Ph.D, Professor of Statistics, Korea University, Korea |
| Time/Place: |
11:00 - 12:00
FSC1217
|
| | |
| Abstract: |
The trend of massive data use—data mining, data warehousing and
CRM—is one of the most important characteristics in the information
society. Need to combine the customer data with official data
to enhance the value for marketing and business decisions create
a new and increased demand for micro data from the government
agencies. The issue of privacy becomes also important in this
respect when the micro data of such nature are used for private
purposes. One solution to this problem is to design census-based
clusters with very fine segmentations based on demographic similarity
of neighborhood.
Using the Korean Census data we will show the method of combining
data of many different sources, different levels of abstraction
as well as seemingly unrelated types of data. Examples will be
given as to how to improve the data value using the cases of
a few Korean companies.
|
|
|
|
|
