On the Dangers of Cross-Validation. An Experimental Evaluation
R. Bharat Rao IKM CKS Siemens Medical Solutions USA Glenn Fung IKM CKS Siemens Medical Solutions USA
Romer Rosales IKM CKS Siemens Medical Solutions USA
Abstract Cross validation allows models
Stats 241: Data-driven Financial and Risk
Econometrics
Ka Wai Tsang
November 13, 2015
Ka Wai Tsang
Stats 241
Introduction to Survival Analysis
Let be the failure time of an individual from a homogeneous
population. The survival function of is
S(t) = P( >
Stats 241: Data-driven Financial and Risk
Econometrics
Ka Wai Tsang
November 20, 2015
Ka Wai Tsang
Stats 241
Semiparametric approach and partial likelihood
Cox (1972, 1975) introduced a semiparametric method to estimate
the nite-dimensional parameter in t
Stats 241: Data-driven Financial and Risk
Econometrics
Ka Wai Tsang
October 30, 2015
Ka Wai Tsang
Stats 241
Project
Price an Asian call option with payo (S K )+ , with the mean
= n Sit /n computed over n dates spaced t = T /n time units
S
i=1
apart. Assu
Stats 241: Data-driven Financial and Risk
Econometrics
Ka Wai Tsang
October 16, 2015
Ka Wai Tsang
Stats 241
Risk-neutral probability measure
We want to nd a measure Q such that St = e r (T t) E Q (ST | Ft ),
i.e. e rt St = E Q (e rT ST | Ft ), which means
Stats 241: Data-driven Financial and Risk
Econometrics
Ka Wai Tsang
November 6, 2015
Ka Wai Tsang
Stats 241
Importance sampling
Suppose we want to estimate = Ef [h(X )] where f is the density
function of X . Let g be another density that g (x) = 0 wheneve
Stats 241: Data-driven Financial and Risk
Econometrics
Ka Wai Tsang
October 23, 2015
Ka Wai Tsang
Stats 241
Likelihood ratio method
Suppose Y = f (X ) and the density function of X is g (x), and thus
E (Y ) =
f (x)g (x)dx.
If f (x) is continuous almost fo
Stats 241: Data-driven Financial and Risk
Econometrics
Ka Wai Tsang
October 2, 2015
Ka Wai Tsang
Stats 241
Multivariate adaptive regression spline (MARS)
Let Tj denote the set of observed values x1j , . . . , xnj of the jth input
variable.
starts by inclu
Stats 241: Data-driven Financial and Risk
Econometrics
Ka Wai Tsang
October 9, 2015
Ka Wai Tsang
Stats 241
GARCH (Sect. 6.3.2)
Recall an AR(p) for cfw_xt is
xt = + 1 xt1 + + p xtp + ut , E (ut ) = 0.
GARCH(1,1):
2
2
2
ut = t t , t = + t1 + ut1 , E (t ) =
STATS 241 Project
Due Date: Friday, Dec 4, 2015
If you are interested in another question, you can work on your own project but please
send me an email to briey describe your problem by Nov 13.
You may team up with the others to form a group of no more
5601 Notes: Smoothing
Charles J. Geyer
April 8, 2006
Contents
1 Web Pages
2
2 The General Smoothing Problem
2
3 Some Smoothers
3.1 Running Mean Smoother . .
3.2 General Kernel Smoothing . .
3.3 Local Polynomial Smoothing
3.4 Smoothing Splines . . . . . .
Vol. 51, No. 11, November 2005, pp. 16431656 issn 0025-1909 eissn 1526-5501 05 5111 1643
MANAGEMENT SCIENCE
informs
doi 10.1287/mnsc.1050.0415 2005 INFORMS
Importance Sampling for Portfolio Credit Risk
Columbia Business School, Columbia University, New Yo
STAT241 ASSIGNMENT 1
DUE 4/24/2009
Part I: Option Pricing and Nonparametric Regression. Syllabus: Sections 7.1-7.4, 8.1-8.3 For Stanford Student
Problems: 7.3, 7.5, 8.2, 8.3 Instructions on how to hand in the homework: Only hard copy accepted. Either in c
STAT241 ASSIGNMENT 2
DUE 5/15/2009
Part II: Interest Rate model and Advanced Time Series Syllabus: Chapter 9 and 10 Paul Pong(ckpong@stanford.edu) will hold oce hour and grade this assignment. His oce hour is 2:30-3:30pm Tuesday and Thursday in Room 108.
STAT241 ASSIGNMENT 3
DUE 5/29/2009
Part III: Statistical Trading Strategies and Risk Management Syllabus: Chapter 11 and 12 Kevin Wu(kevin.bwu@stanford.edu) will hold oce hour and grade this assignment. His oce hour is 2:30-3:30pm Tuesday and Thursday in
STAT241 PROJECT
DUE 6/8/2009
1 Interest rate models: Problem 9.5 and 10.5 Before carrying out problem 9.5 (a) (b) (c) (d), you should plot the time series of three rates and their autocorrelation functions, and also, examine corresponding plots for the di
Model Robust Regression
Based on Generalized Estimating Equations
by
Seth K. Clark
Dissertation submitted to the faculty of Virginia
Polytechnic Institute and State University in partial
fulﬁllment of the requirements for the degree of
Doctor of Philosoph
Stats 241: Data-driven Financial and Risk
Econometrics
Ka Wai Tsang
September 25, 2015
Ka Wai Tsang
Stats 241
Instructor
Ka Wai Tsang (ktsang@stanford.edu)
O ce Hour: Rm 220, Sequoia, Thurs 5-6 pm
TAs
Wenfei Du (wdu@stanford.edu)
Jiyao Kou (jiyaokou@stanf