2011-2012 Bulletin â PDF - SEAS Bulletin - Columbia University
2011-2012 Bulletin â PDF - SEAS Bulletin - Columbia University
2011-2012 Bulletin â PDF - SEAS Bulletin - Columbia University
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
204<br />
STAT W4240x Data mining<br />
3 pts. Professors Madigan and Wood.<br />
Prerequisite: COMS W1003, W1004, W1005,<br />
W1007, or the equivalent. Corequisites: Either<br />
STAT W3105 or W4105, and either STAT W3107<br />
or W4107. Data Mining is a dynamic and fast<br />
growing field at the interface of Statistics and<br />
Computer Science. The emergence of massive<br />
datasets containing millions or even billions of<br />
observations provides the primary impetus for the<br />
field. Such datasets arise, for instance, in largescale<br />
retailing, telecommunications, astronomy,<br />
computational and statistical challenges. This<br />
course will provide an overview of current<br />
research in data mining and will be suitable for<br />
graduate students from many disciplines. Specific<br />
topics covered with include databases and data<br />
warehousing, exploratory data analysis and<br />
visualization, descriptive modeling, predictive<br />
modeling, pattern and rule discovery, text mining,<br />
Bayesian data mining, and causal inference.<br />
STAT W4290y Statistical methods in finance<br />
3 pts. Professor Pospisil.<br />
Prerequisite: STAT W3107 or W4107. A fastpaced<br />
introduction to statistical methods used in<br />
quantitative finance. Financial applications and<br />
statistical methodologies are intertwined in all<br />
lectures. Topics include regression analysis and<br />
applications to the Capital Asset Pricing Model and<br />
multifactor pricing models, principal components<br />
and multivariate analysis, smoothing techniques<br />
and estimation of yield curves statistical methods<br />
for financial time series, value at risk, term<br />
structure models and fixed income research, and<br />
estimation and modeling of volatilities. Hands-on<br />
experience with financial data.<br />
STAT W4315x and y Linear regression models<br />
3 pts. Instructor to be announced.<br />
Prerequisites: STAT W3107 or equivalent, MATH<br />
V2110 or the equivalent. Corequisites: MATH<br />
V1101, V1102, and V2110. Simple and multiple<br />
regression, including testing, estimation and<br />
confidence procedures, modeling, regression<br />
diagnostics and plots, polynomial regression, fixed<br />
effects ANOVA and ANCOVA models, nonlinear<br />
regression, multiple comparisons, colinearity<br />
and confounding, model selection. Emphasis on<br />
geometric approach to the theory and the use of a<br />
statistical package to analyze data.<br />
STAT W4325y Generalized linear models<br />
3 pts. Professor Sobel.<br />
Prerequisite: STAT W4315. Statistical methods<br />
for rates and proportions, ordered and nominal<br />
categorical responses, contingency tables, oddsratios,<br />
exact inference, logistic regression, Poisson<br />
regression, generalized linear models.<br />
STAT W4330x Multilevel models<br />
Professor Chen.<br />
Prerequisites: STAT W4315. Theory and practice,<br />
including model-checking, for random and mixedeffects<br />
models (also called hierarchical, multilevel<br />
models). Extensive use of the computer to<br />
analyze data.<br />
STAT W4335x Sample surveys<br />
3 pts. Professor Sobel.<br />
Prerequisite: STAT W3107 or W4107. Introductory<br />
course on the design and analysis of sample<br />
surveys. How sample surveys are conducted,<br />
why the designs are used, how to analyze survey<br />
results, and how to derive from first principles<br />
the standard results and their generalizations.<br />
Examples from public health, social work, opinion<br />
polling, and other topics of interest.<br />
STAT W4413y Nonparametric statistics<br />
3 pts. Professor Sen.<br />
Prerequisite: STAT W3107 or W4107. Statistical<br />
inference without parametric model assumption.<br />
Hypothesis testing using ranks, permutations, and<br />
order statistics. Nonparametric analogs of analysis<br />
of variance. Non-parametric regression, smoothing<br />
and model selection.<br />
STAT W4437x and y Time series analysis<br />
3 pts. Professors Davis, Hernandez-del-Valle,<br />
and Hueter.<br />
Prerequisite: STAT W4315 or equivalent. Least<br />
squares smoothing and prediction, linear systems,<br />
Fourier analysis, and spectral estimation. Impulse<br />
response and transfer function. Fourier series, the<br />
fast Fourier transform, autocorrelation function, and<br />
spectral density. Univariate Box-Jenkins modeling<br />
and forecasting. Emphasis on applications.<br />
Examples from the physical sciences, social<br />
sciences, and business. Computing is an integral<br />
part of the course.<br />
STAT W4543y Survival analysis<br />
Professor Shnaidman.<br />
Prerequisite: STAT W4315. Survival distributions,<br />
types of censored data, estimation for various<br />
survival models, nonparametric estimation of<br />
survival distributions, the proportional hazard<br />
and accelerated lifetime models for regression<br />
analysis with failure-time data. Extensive use of<br />
the computer.<br />
STAT W4606x and y Elementary stochastic<br />
processes<br />
3 pts. Professors Brown and Hogan.<br />
Prerequisite: STAT W3105, W4105, or equivalent.<br />
Review of elements of probability theory. Poisson<br />
processes. Renewal theory. Wald’s equation.<br />
Introduction to discrete and continuous time<br />
Markov chains. Applications to queueing theory,<br />
inventory models, branching processes.<br />
STAT W4635y Stochastic Processes for<br />
Finance<br />
3 pts. Professor Vecer.<br />
Prerequisite: STAT W3105, W4105, or equivalent.<br />
This course covers theory of stochastic<br />
processes applied to finance. It covers concepts<br />
of Martingales, Markov chain models, Brownian<br />
motion. Stochastic Integration, Ito’s formula<br />
as a theoretical foundation of processes used<br />
in financial modeling. It also introduces basic<br />
discrete and continuous time models of asset price<br />
evolutions in the context of the following problems<br />
in finance: portfolio optimization, option pricing,<br />
spot rate interest modeling.<br />
STAT W4840x Theory of interest<br />
3 pts. Professor Rajah.<br />
Prerequisite: MATH V1101 or equivalent.<br />
Introduction to the mathematical theory of interest<br />
as well as the elements of economic and financial<br />
theory of interest. Topics include rates of interest<br />
and discount; simple, compound, real, nominal,<br />
effective, dollar (time)-weighted; present, current,<br />
future value; discount function; annuities; stocks<br />
and other instruments; definitions of key terms<br />
of modern financial analysis; yield curves; spot<br />
(forward) rates; duration; immunization; and short<br />
sales. The course will cover determining equivalent<br />
measures of interest; discounting; accumulating;<br />
determining yield rates; and amortization.<br />
engineering <strong>2011</strong>–<strong>2012</strong>