class proposal by Yannis

Hi Vince,
Yannis of the Weather desk is planning to develop relationship with Prof
Rene Carmona in doing weather analysis. To start this off, they are planning
to pay Prof Carmona to give a training class as outlined below, and they
want to know if Research is willing to send people and bear part of the costs.

We can talk more at 4:00 pm, but while there is no doubt that getting people
from outside to present new ideas is always important and interesting, I
think Research Group members can easily give most of these talks.
Apparently, people are interested in these topics and are willing to pay to
listen. My thought is that if the intent is to develop relationships, that
is fine, but the Research Group should also be given the opportunity to
provide more training and get more visibility. I have already communicated
this to Joe and to Yannis.

Vasant

---------------------- Forwarded by Vasant Shanbhogue/HOU/ECT on 03/05/2001
10:40 AM ---------------------------
From: Yannis Tzamouranis/ENRON@enronXgate on 03/05/2001 10:01 AM
To: Vasant Shanbhogue/HOU/ECT@ECT
cc:
Subject:

Here is a tentative course description.

Day 1: Extreme Value Distriutions and Copulas
1. Heavy Tail Distributions: Exploratory data analysis and detection.
Extreme value distributions and generalized Pareto distributions.
Estimation and simulation. Practical examples.
2. Notions of dependence and copulas. Estimation and simulation.
Experiments with the program EVANECE.

Day 2: Principal Component Analysis and Modern Regression
1. Principal component analysis and applications to the yield curve and
the detection of contagion in financial markets.
2. Nonlinear regression and the construction of yield curves.
3. Nonparametric regression (kernel and projection pursuit methods) and
alternatives to the Black-Scholes formula to option pricing.

Day 3: Time Series Analysis
Examples of temperature time series will be used to introduce and
illustrate the following concepts and techniques:
1. Removing trends and seasonal components, and stationarity.
2. Fitting the classical autoregressive and moving average models.
3. Discretization of stochastic differential equations
4. Multivariate time series

Day 4: Nonlinear Systems and Filtering
1. ARCH, GARCH and stochastic volatility models
2. Linear state space models and the classical Kalman filter
3. Nonlinear systems and particle filtering.
4. Applications