Ryo
Mizuno
Master's Degree Plan B Presentation:
Tues., December 1, 2009, 1:30, Room TBA
Ryo Mizuno
UHM Math. Depart.
Core and No-Treat Equilibrium in Tournament Games with Externalities
talks/ryo_mizuno_master_talk.pdf
We consider a situation where coalitions are formed
to divide a resource. As in real life, the value of a payoff to a given player
is allowed to depend on the payoff to other players with whom he shares a common
interest. There are various notions of equilibrium for this type of game,
including the core and no-treat equilibrium. These stabilities may exist or not,
depending on the power structure and the rule for allocating the resource. It is
shown that under certain conditions, the no-treat equilibrium can exist even
though the core is empty.
QuinnCulver
Master's Degree Plan B Presentation:
Tues., June 15, 2010, 1:30, Keller 401Quinn CulverUHM Math. Depart.Polynomial-clone reducibility
talks/Culver_masters_talk_Polynomial-clone_Reducibilities.pdf
Polynomial-clone reducibility is a generalization of
truth-table reducibility. A polynomial clone is a set of functions over a finite
set X that is closed under composition and contains the constant and projection
functions.
MichelleManes
Honors Seminar for NSF Scholars:
Tues., Sep. 9, 2008, 1:30, Keller 303 Prof Michelle ManesUHM Math Dept
talks/manes_benford_law.pdf
Benford's Law: Tables of Logarithms, Tax Cheats, and The Leading Digit Phenomenon
GalenSasaki
Honors Seminar for NSF Scholars:
Tues., Sep. 16, 2008, 1:30, Keller 303 Prof Galen SasakiUHM Electrical Engineering DeptCrossbars are Fun!
talks/sasaki.pdf>
DanielJenkins
Honors Seminar for NSF Scholars:
Tues., Sep. 23, 2008, 1:30, Keller 303 Prof. Daniel JenkinsUHM Molecular Biosciences and Bioengineering
Engineering a disposable platform for DNA detection
NeilBerg
Honors Seminar for NSF Scholars:
Tues., Sep. 30, 2008, 1:30, Keller 303 Dr. Neil BergUH Consulting HydrologistThe Arrowhead Tunnels ProjectJuliaPatriarche
Honors Seminar for NSF Scholars:
Tues., Nov. 25, 2008, 1:30, Keller 303 Prof Julia PatriarcheUHM ICS DepartmentComputers in Medicine: Change Detection in <br />
Serial Magnetic Resonance Imaging Studies of Brain Tumor Patients
IanRobertson
Honors Seminar for NSF Scholars:
Tues., Oct. 28, 2008, 1:30, Keller 303 Prof Ian RobertsonUHM Dept of Civil EngineeringDevelopment of Performance-based Tsunami EngineeringRudyTamayo
Honors Seminar for NSF Scholars:
Tues., Oct. 21, 1:30, Keller 303 Rudy TamayoHawaiian Electric Co.Power EngineeringWesPeterson
Honors Seminar for NSF Scholars:
Tues., Oct. 7, 2008, 1:30, Keller 303 Prof Wes PetersonUHM ICS DeptAlgorithms for election security
Vu
Nguyen
Master's Degree Plan B Presentation
Wed., July 1, 2009, 2:00, Keller 414
Vu Nguyen
UHM Math. Depart.
OMEGA FUNCTION: A THEORETICAL INTRODUCTION
talks/Nguyen_masters_talk.pdf
The purpose of this paper is to investigate the
theory behind the new universal performance measure (the so called Omega
function), which was first introduced by Con Keating and William F. Shadwick in
2002.
In the first section, we review some rudimentary probability. We then define the
Omega function, introduce some of its properties, and prove these properties
without continuity assumptions.
We also define the standard dispersion, a new statistic derived from the Omega
function. We prove one new theorem about the range of the standard dispersion
for a finite sample. The structure of the second section on the Omega function
follows closely that of a recent talk given by Ana Cascon and William Shadwick.
In the last section, we demonstrate these properties with real-life data.
Chee
Chen
Master's Degree Plan B Presentation
Fri., May 1, 2009, 3:00, Keller 402
Chee Chen
UHM Math. Depart.
THE SHANGHAI STOCK EXCHANGE:
STATISTICAL PROPERTIES AND SIMULATION
talks/XiongzhiChen_masters_talk.pdf
This thesis contains two independent papers that summarize the author's study on
the statistical properties of the returns series of the daily price index series
of the Shanghai Stock Exchange (SSE) (hereunder called "returns series" ) for
trading days between 01/01/1991 and 14/11/2007, and the simulation of the
detrended logarithmic SSE index series (hereunder called "detrended series" ) by
the ergodic, generalized hyperbolic (GH) diffusion whose invariant density is
the generalized hyperbolic (GH) density.
More specifically, the first paper herein gives the right tail index estimate,
the left tail index estimate, for the returns series, and the Hurst index for
the absolute returns series, all of which are consistent with what most
researchers obtained for stock market index series and tend to argue that
extreme care should be taken when making assumptions on the underlying
stochastic process that generates the stock price index series. These estimates
are obtained by the simplest method -- the traditional ordinary least squares
(OLS) method, to avoid unnecessary or even wrong assumptions on stationarity or
dependence structures of the index series, whereas some comparative estimates
via different methods are also provided.
The second paper is devoted to testing out the ergodic GH diffusion by the
detrended series. It constructs for the detrended series an ergodic GH diffusion
whose invariant density is proportional to the GH density and estimates the
unknown parameter vector by Markov chain Monte Carlo (MCMC) methods, which
induces six possible models from six independent, convergent Markov chains by
the Monte Carlo standard error (MCSE) convergence diagnostic. Unfortunately the
Gelman-Rubin convergence diagnostic failed to converge to one even after 350,000
iterations of the Metroplolis-Hastings algorithm, since the six chains converged
to their own stationary distributions at around 50,000 iterations.
After six Monte Carlo estimates of unknown parameter vector are obtained,
uniform residual test is carried out for four of the six possible models due to
the computational capability of Mathematica and Matlab, and the test results
statistically reject all four selected models, thus rejecting the null
hypothesis that the underlying stochastic process for the detrended index is an
ergodic GH diffusion.