Similar documents
A L A BA M A L A W R E V IE W

Topic 7 - Matrix Approach to Simple Linear Regression. Outline. Matrix. Matrix. Review of Matrices. Regression model in matrix form

Scripture quotations marked cev are from the Contemporary English Version, Copyright 1991, 1992, 1995 by American Bible Society. Used by permission.

Likelihood Inference for Lattice Spatial Processes

/ / MET Day 000 NC1^ INRTL MNVR I E E PRE SLEEP K PRE SLEEP R E

Infinitely Imbalanced Logistic Regression

176 5 t h Fl oo r. 337 P o ly me r Ma te ri al s

22s:152 Applied Linear Regression. Example: Study on lead levels in children. Ch. 14 (sec. 1) and Ch. 15 (sec. 1 & 4): Logistic Regression

Generalized Linear Models. Kurt Hornik

Final Examination. Tuesday December 15, :30 am 12:30 pm. particles that are in the same spin state 1 2, + 1 2

Link lecture - Lagrange Multipliers

ECE531 Homework Assignment Number 7 Solution

An Efficient Estimation Method for Longitudinal Surveys with Monotone Missing Data

Fun and Fascinating Bible Reference for Kids Ages 8 to 12. starts on page 3! starts on page 163!

i=1 k i=1 g i (Y )] = k

MLE for a logit model

S U E K E AY S S H A R O N T IM B E R W IN D M A R T Z -PA U L L IN. Carlisle Franklin Springboro. Clearcreek TWP. Middletown. Turtlecreek TWP.

(a) To show f(z) is analytic, explicitly evaluate partials,, etc. and show. = 0. To find v, integrate u = v to get v = dy u =

Central Limit Theorem ( 5.3)

Chapter 5 Matrix Approach to Simple Linear Regression

CMO 2007 Solutions. 3. Suppose that f is a real-valued function for which. f(xy) + f(y x) f(y + x) for all real numbers x and y.

Math 3330: Solution to midterm Exam

Logistic regression. 11 Nov Logistic regression (EPFL) Applied Statistics 11 Nov / 20

Chapter 17 The bilinear covariant fields of the Dirac electron. from my book: Understanding Relativistic Quantum Field Theory.

MATH 644: Regression Analysis Methods

Ph.D. Qualifying Exam Friday Saturday, January 6 7, 2017

VERITAS L1 trigger Constant Fraction Discriminator. Vladimir Vassiliev Jeremy Smith David Kieda

Controllability of linear PDEs (I): The wave equation

Inference After Variable Selection

Gauge Plots. Gauge Plots JAPANESE BEETLE DATA MAXIMUM LIKELIHOOD FOR SPATIALLY CORRELATED DISCRETE DATA JAPANESE BEETLE DATA

Approximation of the Fisher information and design in nonlinear mixed effects models

ME 509, Spring 2016, Final Exam, Solutions

x 3y 2z = 6 1.2) 2x 4y 3z = 8 3x + 6y + 8z = 5 x + 3y 2z + 5t = 4 1.5) 2x + 8y z + 9t = 9 3x + 5y 12z + 17t = 7

Exercise 9: Model of a Submarine

Single-level Models for Binary Responses

Structure of Generalized Parton Distributions

Last Lecture - Key Questions. Biostatistics Statistical Inference Lecture 03. Minimal Sufficient Statistics

Solutions to Assignment #8 Math 501 1, Spring 2006 University of Utah

EE5120 Linear Algebra: Tutorial 7, July-Dec Covers sec 5.3 (only powers of a matrix part), 5.5,5.6 of GS

[ zd z zdz dψ + 2i. 2 e i ψ 2 dz. (σ 2 ) 2 +(σ 3 ) 2 = (1+ z 2 ) 2

Lecture 2: Martingale theory for univariate survival analysis

2.7 Estimation with linear Restriction

8 Nonlinear Regression

The Schroedinger equation

I. Relationship with previous work

LIST OF FORMULAS FOR STK1100 AND STK1110

Theoretical Tutorial Session 2

Parts Manual. EPIC II Critical Care Bed REF 2031

FORMULAE FOR G-FUNCTION

3.6. Let s write out the sample space for this random experiment:

APPLICATIONS OF FD APPROXIMATIONS FOR SOLVING ORDINARY DIFFERENTIAL EQUATIONS

F l a s h-b a s e d S S D s i n E n t e r p r i s e F l a s h-b a s e d S S D s ( S o-s ltiad t e D r i v e s ) a r e b e c o m i n g a n a t t r a c

At point G V = = = = = = RB B B. IN RB f

Solutions to PS 2 Physics 201

Chemistry 532 Problem Set 7 Spring 2012 Solutions

Econometrics II - EXAM Answer each question in separate sheets in three hours

H U G H T H O M A S, C E O, S K Y H I G H W A Y S

Econometrics A. Simple linear model (2) Keio University, Faculty of Economics. Simon Clinet (Keio University) Econometrics A October 16, / 11

Fuzzy Reasoning and Optimization Based on a Generalized Bayesian Network

Data Mining Stat 588

Solutions to Exercises 6.1

Estimating the Marginal Odds Ratio in Observational Studies

Qualification Exam: Mathematical Methods

Econometric modelling and forecasting of intraday electricity prices

Lecture 04. Curl and Divergence

IMPORTANCE SAMPLING Importance Sampling Background: let x = (x 1,..., x n ), θ = E[h(X)] = h(x)f(x)dx 1 N. h(x i ) = Θ,

Combining Non-probability and Probability Survey Samples Through Mass Imputation

OH BOY! Story. N a r r a t iv e a n d o bj e c t s th ea t e r Fo r a l l a g e s, fr o m th e a ge of 9

Differential forms. Proposition 3 Let X be a Riemann surface, a X and (U, z = x + iy) a coordinate neighborhood of a.

p y (1 p) 1 y, y = 0, 1 p Y (y p) = 0, otherwise.

Economics 620, Lecture 5: exp

Nonparametric estimation of extreme risks from heavy-tailed distributions

Gibbs models estimation of galaxy point processes with the ABC Shadow algorithm

Solutions to exam : 1FA352 Quantum Mechanics 10 hp 1

Regression Analysis Chapter 2 Simple Linear Regression

Chapter 5: Models used in conjunction with sampling. J. Kim, W. Fuller (ISU) Chapter 5: Models used in conjunction with sampling 1 / 70

Mathematics for 3D Graphics

Gifted Education Program Plan

Conditional Distributions

F & B Approaches to a simple model

Quantitative Methods in Economics Panel data and generalized least squares

BIOS 2083 Linear Models c Abdus S. Wahed

A = (a + 1) 2 = a 2 + 2a + 1

c. What is the average rate of change of f on the interval [, ]? Answer: d. What is a local minimum value of f? Answer: 5 e. On what interval(s) is f

EE/Stat 376B Handout #5 Network Information Theory October, 14, Homework Set #2 Solutions

University of Oxford. Statistical Methods Autocorrelation. Identification and Estimation

p r * < & *'& ' 6 y S & S f \ ) <» d «~ * c t U * p c ^ 6 *

Contents. MATH 32B-2 (18W) (L) G. Liu / (TA) A. Zhou Calculus of Several Variables. 1 Multiple Integrals 3. 2 Vector Fields 9

Jesper Møller ) and Kateřina Helisová )

SGN Advanced Signal Processing: Lecture 8 Parameter estimation for AR and MA models. Model order selection

El X Cs. Allen Percival. a I i I ( INSTRUCTOR(S) Training Roster TRAINING TOPIC DATE/TIME OFTRAINING. PRINT NAME (Legibly)

Fall 2017 STAT 532 Homework Peter Hoff. 1. Let P be a probability measure on a collection of sets A.

Bayesian Interpretations of Regularization

Econ 5150: Applied Econometrics Dynamic Demand Model Model Selection. Sung Y. Park CUHK

Boltzmann Equation and Hydrodynamics beyond Navier-Stokes

IP WEIGHTING AND MARGINAL STRUCTURAL MODELS (CHAPTER 12) BIOS IPW and MSM

(308 ) EXAMPLES. 1. FIND the quotient and remainder when. II. 1. Find a root of the equation x* = +J Find a root of the equation x 6 = ^ - 1.

Generalized linear models

G-ESTIMATION OF STRUCTURAL NESTED MODELS (CHAPTER 14) BIOS G-Estimation

Physics 7A Lecture 2 Fall 2014 Final Solutions. December 22, 2014

Transcription:

8.1 8.2 9.1 9.2 9.3 10.1 10.2 11.1 12.1 12.2 13.1 14.2 15.1 15.2 16.1 17.1 17.2 8.1 8.2 9.1

9.2 10.1 10.2 10.3 11.1 12.1 12.2 13.1 13.2 14.1 14.1 14.2 15.1 16.1 16.2 16.3 16.4 16.5 16.6

W A W C 12

U A 7 8

U A U Y U A U Y

16

95 95 95

1 i p a=1 i 2 Y i = Yi a=1 p a=1 i 3 p a=1 i p 13 2 1 10.1 10.2 U L L L A L U A L L A

ˆθ 0 θ 1 W ˆθ 0 ˆθ 1 W i [Y i θ 0 + θ 1 A i ] 2. [Y A = a] = θ 0 a A = a i i=1 Y iw i i=1 W i 95 θ 1 = 3.4 W A = 1/f(A L) SW A = f(a)/f(a L) 95

a E[Y a ] = β 0 + β 1 a. V E[Y a A, V ] = β 0 + β 1 a + β 2 V a + β 3 V. β 3 V A V L A B A B A B

W A,C = W A W C A, C L Y a=1,c=0 Π(A, C) L L SW A,C = SW A SW C SW C = P r[c = 0 A]/P r[c = 0 L, A] L C L L(and ) W A,C = 1/f(A, C = 0 L) A C f(a, C = 0 L) = f(a L) P r[c = 0 L, A] L f(a L) P r[c = 0 L, A] L L C A L E[Y a=1,c=0 ]

E[Y a=1,c=0 ] A = 1 A = 0 a E[Y A = a, C = 0, L = l] P r[l = l]. l L P r[l = l] P DF f L (l) Y E[Y A = a, C = 0, L = l]df L (l). A 9 L L l E[Y A = a, C = 0, L = l] A

l E[Y A = a, L = l] P r[l = l] A = 1 A = 0 A Y L L

l L L L A SW A (L) 1 L P r[a = 1 Y a=0, L] = P r[a = 1 L]

A logitp r[a = 1 Y a=0, L] = α 0 + α 1 Y a=0 + α 2 L α 2 L L p L 1,..., L p α 2 L = p j=1 α 2jL j α 1 Y a=0 A L E[Y a Y a=0 A = a, L] = β 1 a + β 2 al L β 0 β 3 L Y a=1 Y a=0

A Y A Y L Y a i Y a=0 i = ϕ 1 a + ϕ 2 al i ϕ 1 + ϕ 2 l L = l Y a=0 Y a=1 L E[Y a Y a=0 A = a, L] = β 1 a Yi a Yi a=0 = ϕ 1 a i ϕ 1 β i Y a Y a=0 = ϕ 1 a Y a=0 = Y a ϕ 1 a Y a=0 = Y ϕ 1 a ϕ 1 Y a=0 ϕ 1

ϕ 1 H(β 1 ) = Y a=1 H(β 1 ) Y a=0 L V V V E[Y a Y a=0 A = a, L] = β 1 a + β 2 av Y a i Y a=0 i = ϕ 1 a + ϕ 2 av logitp r[a = 1 H(ϕ ), L] = α 0 + α 1 H(ϕ ) + α 2 H(ϕ )V + α 3 L. A

L Y A E[Y a,c=0 L] = β 0 + β 1 a + β 2 al + β 3 L E[Y a,c=0 L = l] E[Y A = 1, C = 0, L = l] E[Y A = 1, C = 0, L = l] = α 0 + α 1 a + α 2 al + α 3 L L L p(l) Y a A L Y a A p(l) p(l) 1 0 L p(l) p(l)

E[Y A, C = 0, p(l)] s E[Y a=1,c=0 p(l) = s] E[Y a=0,c=0 p(l) = s] E[Y a=1,c=0 ] E[Y a=0,c=0 ] E[Y A, C = 0, p(l)] 13 L p(l) p(l) s p(l) s ± 0.05 A L L A L A A Y a L a

Z A Z Y A Z Y E[Y a=1 ] E[Y a=0 ] E[Y Z = 1] E[Y Z = 0] E[A Z = 1] E[A Z = 0] Z Cov(Y,Z) Cov(A,Z) Z Y Z A

E[A Z] = α 0 + α 1 Z Ê[A Z] E[Y Z] = β 0 + β 1 Ê[A Z] ˆβ 1 10 A Y Z E[Y a=1 Y a=0 Z = 1, A = a] = E[Y a=1 Y a=0 Z = 0, A = a] a = 0, 1 A z=1 = 1 A z=0 = 1 A z=1 = 0 A z=0 = 0 A z=1 = 1 A z=0 = 0 A z=1 = 0 A z=0 = 1

A z=1 A z=0 E[Y a=1 Y a=0 A z=1 = 1, A z=0 = 0]. Z Y Z Z Y Z A A Z Y Z 16.2 16.3 U z U z A uz u z U z

Z A Z A A U A U Z A Z 95 A A Z Y Z Y V

Z Z Y

k P r[t > k] k k 1 P r[t > k] = P r[t k] t k t P r[t = k T > k 1] k k k k t k

D k k P r[d k = 0] = P r[t > k] k P r[d k = 1] = P r[t k] k P r[d k = 1 D k 1 = 0] k = 1 k = 0 k 0 k P r[d k = 0] = k P r[d m = 0 D m 1 = 0] m=1 k k P r[d k = 1 D k 1 = 0] k k 1 P r[d k = 0] k

k P r[d k+1 = 1 D k = 0] logitp r[d k+1 = 1 D k = 1, A] = θ 0,k + θ 1 A + θ 2 A k + θ 3 A k 2 θ 0,k = θ 0 + θ 4 k + θ 3 A k 2 P r[d k+1 = 0 A = a] P r[d k+1 = 0 D k = 0, A = a]

8.1 8.2 A = 1 E = 1 A = 1 E = 1 9.1 A Y U A A U Y Y

9.2 9.3 10.1 95 n 95 n

10.2 11.1 95 95 95 95 95 95$ 95 12.1 12.2

0 13.1 P r[a = a L = l] = 0 P r[l = l] 0 E[Y A = a, L = l] E[Y A, L] 95 95 14.2 α 1 = 0 L

α 1 15.1 E[Y a,c=0 L] = β 0 + β 1 a + β 2 al + β 3 L, β 1 β 2 β 0 β 3 15.2 16.1 Z A Y A

Z A Y Z A Y 17.1 72 17.2

8.1 8.2 P r[y = 0 E = e, A = a] P r[y = 0 E = e, A = a] = g(e)h(a) P r[y = 0 E = e, A = a]/p r[y = 0 E = e, A = 0] P r[y = 0 E = e, A = a] = g(e)h(a) P r[y = 1 E = e, A = a] = 1 g(e)h(a) Y = 0 Y = 1 9.1 f( ) P DF U A U Y P DF P DF U A P DF U Y P DF 9.2 Y z=0,a = Y z=1,a

10.1 10.2 Y A L A L L 10.3

11.1 p i=0 θ ix i g{ } g{e[y X]} = p i=1 θ ix i log{e[y X]} = p θ i X i. i=1 E[Y X] p log{ 1 E[Y X] } = θ i X i. i=1 θ E[Y X] n i=1 E[Y X = x] x ω h(x X i )Y I i=1 nω h(x X i ω ) h (z) z = 0 0 z h ω p i=0 θ ix i p i=0 f i(x i )

f i ( ) k E[Y X = x] E[Y X = x] X x X X x E[Y X = x] E[Y X = x] X x E[Y X = x] X x h h = 0 12.1 I(A = a)y ÊE[ ] f(a L) ÊE[ I(A=a)Y f(a L) ] ÊE[ I(A=a) f(a L) ]

12.2 g[a] f[a L] E[Y a ] E[ I(A = a) E[ ] = 1 f[a Y ] I(A = a)y E[ ] = E[Y a ] f[a Y ] E[ I(A=a)Y ] f[a Y ] E[ I(A=a) ] = E[Y a ] f[a Y ] E[ I(A=a)Y f[a Y ] E[ I(A=a) I(A = a)y f[a Y ] g(a)] f[a Y ] g(a)] = E[Y a ] g(a)] = E[Y a ]g(a) E[ I(A = a) g(a)] f[a Y ] 13.1 95 95 13.2 A Y E[Y A = a, C = 0, L = l, D]

A = 1 A = 0 L 14.1 a V L E[Y a V ] = β 0 + β 1 a + β 2 av + β 3 V. E[Y a V ] = E[Y a=0 V ] + β 1 a + β 2 av E[Y a Y a=0 V ] = β 1 a + β 2 av 14.1 log( E[Y a A = a, L] E[Y a=0 A = a, L] ) = β 1a + β 2 L H(ϕ ) Y exp[ ϕ 1a ϕ 2L] (0, 1) Y = 1 L 1 Y logitp r[y a = 1 A = a, L] logitp t[y a=0 = 1 A = a, L] = β 1 a + β 2 L

14.2 β 1 H(ϕ ) H(ϕ ) A ϕ N i=1 I[C i = 0]W C i H i (ϕ )(A i E[A L i ]) = 0 H i (ϕ ) = Y i ϕ A i ϕ 1 = N i=1 I[C i = 0]W C i Y i (A i E[A L i ])/ N i=1 I[C i = 0]W C i A i (A i E[A L i ]) H i (ϕ ) H(ϕ ) H(ϕ ) E[H(ϕ ) L] ϕ E[H(ϕ ) L] E[A L] P r[c = 1 A, L] 15.1 b(l) L A L b(l) L b(l)

16.1 Z A Z Y Y a,z Z a, z Z 16.2 P r[y a=1 = 1] P r[y a=0 = 1] Z A Y 16.2 16.3 A Z E[Y a=1 Y a=0 A = 1, Z] = β 0 + β 1 Z E[Y Y a=0 A, Z] = A(β 0 + β 1 Z) β 0 Z = 0

β 0 + β 1 Z = 1 Z β 1 = 0 β 0 E[Y a=0 Z = 1] = E[Y a=0 Z = 0] E[Y A(β 0 + β 1 ) Z = 1] = E[Y Aβ 0 Z = 0] β 1 = 0 β 0 = E[Y Z = 1] E[Y Z = 0] E[A Z = 1] E[A Z = 0] β 1 = 0 β 0 = E[Y a=1 Y a=0 A = 1, Z = z] = E[Y a=1 Y a=0 A = 1] z E[Y a=1 ] E[Y a=0 ] β 0 = E[Y a=1 ] E[Y a=0 ] E[Y a=1 ] E[Y a=0 ] 16.4 A Z E[Y a=1 A = 1, Z] E[Y a=0 A = 1, Z] = exp(β 0 + β 1 Z), E[Y A, Z] = E[Y a=0 A, Z]exp[A(β 0 + β 1 Z), exp(β 0 Z = 0 exp(β 0 + β 1 ) Z = 1

β 1 = 0 E[Y a=1 ]/E[Y a=0 ] = exp(β 0 ) E[Y a=1 ] E[Y a=0 ] = E[Y A = 0](1 E[A])[exp(β 0 ) 1] + E[Y A = 1]E[A][1 exp(β 0 )] Z E[Y (1)] E[Y (0)] Z 16.5 E[Y Y a=0 Z, A, V ] = γ(z, A, V, ψ ) E[Y a=0 Z = 1, V ] = E[Y a=0 Z = 0, V ]. E[Y Z, A, V ] = E[Y 0 Z, A, V ]exp[γ(z, A, V, ψ )] 16.6 E[Y a=1 Y a=0 A z=1 A z=0 = 1]

E[Y z=1 Y z=0 ] = E[Y z=1 Y z=0 A z=1 = 1, A z=0 = 1]P r[a z=1 = 1, A z=0 = 1] +E[Y z=1 Y z=0 A z=1 = 0, A z=0 = 0]P r[a z=1 = 0, A z=0 = 0] +E[Y z=1 Y z=0 A z=1 = 1, A z=0 = 0]P r[a z=1 = 1, A z=0 = 0] +E[Y z=1 Y z=0 A z=1 = 0, A z=0 = 1]P r[a z=1 = 0, A z=0 = 1]. E[Y z=1 Y z=0 ] = E[Y z=1 Y z=0 A z=1 = 1, A z=0 = 0]P r[a z=1 = 1, A z=0 = 0]. E[Y z=1 Y z=0 A z=1 = 1, A z=0 = 0] = E[Y a=1 Y a=0 A z=1 = 1, A z=0 = 0]. E[Y a=1 Y a=0 A z=1 = 1, A z=0 = 0] = E[Y z=1 Y z=0 ] P r[a z=1 = 1, A z=0 = 0]. Z E[Y z=1 Y z=0 ] = E[Y Z = 1] E[Y Z = 0]. P r[a z=1 A z=0 = 1] = P r[a = 1 Z = 1] P r[a = 1 Z = 0]. P r[a z=0 = 1] = P r[a = 1 Z = 0], P r[a z=1 = 0] = P r[a = 0 Z = 1]. P r[a z=1 A z=0 = 1] = 1 P r[a = 1 Z = 0] P r[a = 0 Z = 1] = 1 P r[a = 1 Z = 0] (1 P r[a = 1 Z = 1]) = P r[a = 1 Z = 1] P r[a = 1 Z = 0]. 16.1 Z 16.2 16.3 Z U z U z

Z A Y U z U z U z Z U z 10.2

95 11.1 14.2

... L p +E[Y z=1 Y z=0 A z=1 = 1, A z=0 = 0]P r[a z=1 = 1, A z=0 = 0]