Muutke küpsiste eelistusi

E-raamat: Semiparametric Odds Ratio Model and Its Applications

  • Formaat: 306 pages
  • Ilmumisaeg: 19-Dec-2021
  • Kirjastus: CRC Press
  • Keel: eng
  • ISBN-13: 9781351049733
Teised raamatud teemal:
  • Formaat - EPUB+DRM
  • Hind: 67,59 €*
  • * hind on lõplik, st. muud allahindlused enam ei rakendu
  • Lisa ostukorvi
  • Lisa soovinimekirja
  • See e-raamat on mõeldud ainult isiklikuks kasutamiseks. E-raamatuid ei saa tagastada.
Semiparametric Odds Ratio Model and Its ApplicationsSuurem pilt
  • Formaat: 306 pages
  • Ilmumisaeg: 19-Dec-2021
  • Kirjastus: CRC Press
  • Keel: eng
  • ISBN-13: 9781351049733
Teised raamatud teemal:

DRM piirangud

  • Kopeerimine (copy/paste):

    ei ole lubatud

  • Printimine:

    ei ole lubatud

  • Kasutamine:

    Digitaalõiguste kaitse (DRM)
    Kirjastus on väljastanud selle e-raamatu krüpteeritud kujul, mis tähendab, et selle lugemiseks peate installeerima spetsiaalse tarkvara. Samuti peate looma endale  Adobe ID Rohkem infot siin. E-raamatut saab lugeda 1 kasutaja ning alla laadida kuni 6'de seadmesse (kõik autoriseeritud sama Adobe ID-ga).

    Vajalik tarkvara
    Mobiilsetes seadmetes (telefon või tahvelarvuti) lugemiseks peate installeerima selle tasuta rakenduse: PocketBook Reader (iOS / Android)

    PC või Mac seadmes lugemiseks peate installima Adobe Digital Editionsi (Seeon tasuta rakendus spetsiaalselt e-raamatute lugemiseks. Seda ei tohi segamini ajada Adober Reader'iga, mis tõenäoliselt on juba teie arvutisse installeeritud )

    Seda e-raamatut ei saa lugeda Amazon Kindle's. 

Beginning with familiar models and moving onto advanced semiparametric modelling tools Semiparametric Odds Ratio Model and its Applications introduces readers to a new range of flexible statistical models and provides guidance on their application using real data examples. This books range of real-world examples and exploration of common statistical problems makes it an invaluable reference for research professionals and graduate students of biostatistics, statistics, and other quantitative fields.

Key Features:





Introduces flexible statistical models that have yet to systematically introduced in course materials. Discusses applications of the proposed modelling framework in several important statistical problems, ranging from biased sampling designs and missing data, graphical models, survival analysis, Gibbs sampler and model compatibility, and density estimation. Includes real data examples to demonstrate the use of the proposed models, and estimation and inference tools.
Preface xi
1 Odds Ratio Parameter and Its Utilities 1(20)
1.1 Relative risk and odds ratio parameter
1(2)
1.2 Odds ratio parameters in a J x K contingency table
3(3)
1.3 Odds ratio representations of conditional and joint distributions for a J x K contingency table
6(1)
1.4 Maximum likelihood estimators of odds ratio parameters in a J x K contingency table
7(4)
1.5 Link to logit model and logistic regression
11(1)
1.6 Odds ratio parameters in stratified 2 x 2 tables
11(2)
1.7 Common odds ratio parameter
13(2)
1.8 Odds ratio representations for a J x K x M contingency table
15(3)
1.9 Summary and discussion
18(1)
1.10 Exercises
19(2)
2 Odds Ratio Function and Its Modeling 21(34)
2.1 Odds ratio function
21(1)
2.2 Odds ratio decomposition of density functions
22(5)
2.3 Odds ratio representation of densities
27(3)
2.4 Conditional odds ratio function
30(1)
2.5 Odds ratio representation of a joint conditional density
31(1)
2.6 Odds ratio representation of a complex joint density
32(3)
2.7 Relationship between conditional and unconditional odds ratio functions
35(2)
2.8 Hierarchical odds ratio representation for a joint density
37(3)
2.9 Odds ratio functions embedding in a density function
40(3)
2.10 Modeling odd ratio functions in a family of densities
43(3)
2.11 Semiparametric odds ratio model
46(2)
2.12 Extension to relax the positivity condition for the odds ratio representation
48(3)
2.13 Literature on odds ratio functions and relevant statistical models
51(1)
2.14 Exercises
52(3)
3 Estimation and Inference on Semiparametric Odds Ratio Model 55(32)
3.1 An introduction to likelihood-based approaches
55(1)
3.2 Pseudo-likelihood approaches
56(5)
3.2.1 Pairwise and group-wise pseudo-likelihood approaches
56(3)
3.2.2 Asymptotic theory for U-statistics
59(1)
3.2.3 Asymptotic distributions of the pseudo-likelihood estimators
60(1)
3.3 Permutation likelihood approach
61(9)
3.3.1 Approximations using simple Monte Carlo or asymptotics
62(2)
3.3.2 Adaptive Monte Carlo approximation to permutation likelihood
64(2)
3.3.3 Metropolis algorithm for sampling permutations for estimation
66(1)
3.3.4 Permutation likelihood for the joint model
67(3)
3.4 Maximum semiparametric likelihood approach
70(6)
3.4.1 Maximum likelihood estimator for one group of outcomes
70(1)
3.4.2 Computation of the maximum likelihood estimator
71(2)
3.4.3 Maximum likelihood estimator for two groups of outcomes
73(1)
3.4.4 Maximum likelihood for more than two groups of outcomes
74(1)
3.4.5 Large sample behavior of the maximum likelihood estimator
75(1)
3.5 Comparison of different likelihood approaches
76(1)
3.6 The R package SPORM for semiparametric odds ratio model
76(2)
3.7 Simulation study using SPORM
78(4)
3.7.1 Univariate outcome
78(2)
3.7.2 Multivariate outcomes as a group
80(1)
3.7.3 Multivariate outcomes
81(1)
3.8 Data analysis using SPORM
82(2)
3.9 Exercises
84(3)
4 Estimation and Inference on Conditional Odds Ratio Function 87(20)
4.1 A general formulation of the problem
87(1)
4.2 Permutation approach for stratified sample
88(4)
4.2.1 Permutation likelihood approach
88(2)
4.2.2 Application to common odds ratio estimation in stratified 2 x 2 tables
90(2)
4.3 Semiparametric efficient score for estimating conditional odds ratio function
92(3)
4.3.1 Characterization of the nuisance score space and its orthogonal complement
92(1)
4.3.2 The semiparametric efficient score
93(1)
4.3.3 Locally efficient estimator
94(1)
4.4 The special case with one categorical outcome
95(2)
4.5 Doubly robust estimation of conditional odds ratio function
97(5)
4.5.1 An alternative characterization of the orthogonal complement of the nuisance score space
98(1)
4.5.2 The doubly robust property
99(2)
4.5.3 A more specific case
101(1)
4.6 Robust estimation of conditional odds ratio function in multiple outcome models
102(2)
4.7 Summary and literature
104(1)
4.8 Exercises
104(3)
5 Application to Biased Sampling Problems 107(28)
5.1 The general biased sampling problem
107(2)
5.2 Parameter identifiability for outcome-dependent sample
109(6)
5.2.1 Identifiability of model components
109(1)
5.2.2 Parameter identifiability in case-control design
110(2)
5.2.3 Parameter identifiability in outcome-dependent sampling design
112(2)
5.2.4 Extreme-value sampling and length-biased sampling designs
114(1)
5.3 Outcome-dependent sampling designs with covariate matching
115(3)
5.3.1 The general framework
115(1)
5.3.2 Matched case-control design
116(1)
5.3.3 Extreme-value sampling design with matching
117(1)
5.4 Parameter estimation with biased sampling designs
118(3)
5.5 Analysis of misspecified odds ratio model
121(5)
5.5.1 Misspecification in the semiparametric odds ratio model
121(1)
5.5.2 The permutation likelihood approach under misspecified models
121(4)
5.5.3 The maximum semiparametric likelihood approach under misspecified model
125(1)
5.6 Applications
126(6)
5.6.1 Gene-environment independence in case-control genetic association study
126(2)
5.6.2 Secondary traits analysis in a case-control genetic association study
128(2)
5.6.3 Case-only design to study interactions
130(2)
5.7 Summary and literature
132(1)
5.8 Exercises
133(2)
6 Application to Test of Conditional Independence 135(34)
6.1 An introduction to test of conditional independence
135(1)
6.2 Likelihood ratio tests
136(3)
6.2.1 Prospective, retrospective, and joint likelihood ratio tests
136(2)
6.2.2 Likelihood ratio tests based on permutation likelihoods
138(1)
6.3 Likelihood score tests
139(7)
6.3.1 Prospective, retrospective, and joint likelihood score tests
139(5)
6.3.2 Permutation likelihood score tests
144(2)
6.4 Semiparametric efficient score test
146(4)
6.4.1 Semiparametric likelihood formulation
146(2)
6.4.2 Permutation likelihood formulation
148(2)
6.5 Doubly robust test of conditional independence
150(7)
6.5.1 Estimates based on the semiparametric likelihoods
150(3)
6.5.2 Permutation-based doubly robust test of independence
153(3)
6.5.3 Bootstrap implementation of the doubly robust tests
156(1)
6.6 Connections to classical tests of conditional independence
157(3)
6.7 Likelihood ratio test using the R package SPORM
160(2)
6.8 A simulation study of the doubly robust tests
162(3)
6.9 Summary and literature
165(1)
6.10 Exercises
166(3)
7 Application to Network Detection and Estimation 169(22)
7.1 Network and Gaussian graphical model
169(1)
7.2 Gaussian network detection approaches
170(4)
7.2.1 The neighborhood detection approach
170(1)
7.2.2 The joint detection approach
171(1)
7.2.3 The screening approach based on partial correlation coefficient
172(2)
7.3 Network modeling by the semiparametric odds ratio model
174(1)
7.4 Network detection by penalized likelihoods based on permutations
175(6)
7.4.1 Penalized pairwise pseudo-likelihood approach
175(2)
7.4.2 Penalized permutation likelihood approach
177(3)
7.4.3 Hybrid algorithm for increasing the computation speed and estimation efficiency
180(1)
7.5 Network detection by penalized semiparametric likelihood approach
181(3)
7.5.1 Penalized likelihood for neighborhood selection
181(2)
7.5.2 Algorithms for finding the maximum penalized likelihood estimator
183(1)
7.6 Network selection using package SPORM
184(1)
7.7 Software usage cases
185(2)
7.8 Summary and literature
187(1)
7.9 Exercises
188(3)
8 Application to Missing Data Problems 191(26)
8.1 A brief introduction to the missing data problem
191(1)
8.2 The missing covariate problem in regression analysis
192(3)
8.2.1 Missing covariates in parametric regression
192(2)
8.2.2 Missing covariates in the Cox regression model
194(1)
8.3 Monte Carlo algorithm for the likelihood maximization
195(4)
8.3.1 The general formulation
195(2)
8.3.2 Application to parametric regression
197(2)
8.3.3 Application to Cox regression model
199(1)
8.4 Imputation approach to missing covariates
199(9)
8.4.1 A Bayesian framework
199(2)
8.4.2 Draw (γ, G) from the posterior under the consecutive conditional odds ratio models
201(4)
8.4.3 Draw (γp, G) from the posterior under the joint odds ratio model
205(3)
8.4.4 Draw α and θ from their posteriors
208(1)
8.5 Application to nonignorable missing data
208(6)
8.5.1 Nonparametric identifiability of the full data model with missing at random
208(1)
8.5.2 Nonparametric identifiability of the full data model with missing not at random
209(3)
8.5.3 Odd ratio model for the item-wise independent nonresponse data
212(2)
8.6 Summary and literature
214(1)
8.7 Exercises
214(3)
9 Other Applications 217(14)
9.1 Compatibility of conditionally specified models
217(8)
9.1.1 The compatibility problem
217(1)
9.1.2 Compatibility of conditional densities
218(4)
9.1.3 Modifications to incompatible conditionally specified models based on the joint model construction
222(3)
9.1.4 Summary and literature
225(1)
9.2 Semiparametric estimation of multivariate density
225(4)
9.2.1 The semiparametric odds ratio model for density estimation
225(2)
9.2.2 Smoothed estimators for the semiparametric density
227(2)
9.2.3 Summary and literature
229(1)
9.3 Exercises
229(2)
10 Theoretical Results on Estimation and Inference 231(46)
10.1 Variation independence in the joint odds ratio representation of a density
231(1)
10.2 The semiparametric efficient score for semiparametric odds ratio model
232(2)
10.3 Theoretical properties of the semiparametric maximum likelihood estimator
234(9)
10.4 Properties of the maximum penalized semiparametric likelihood estimator
243(25)
10.4.1 Settings for the analysis
243(2)
10.4.2 Theoretical properties of the penalized semiparametric likelihood estimator
245(5)
10.4.3 Supplemental lemmas for the proofs of main results
250(18)
10.5 Bounds for the permutation distribution
268(9)
Bibliography 277(12)
Index 289
Dr. Hua Yun Chen received his PhD in Biostatistics from the University of Michigan. He is currently a Professor of Biostatistics at the University of Illinois at Chicago. His research focuses on statistical methods for incompletely observed data, biased sampling, and epidemiological applications.
Lisainfo e-raamatute kohta Lisainfo e-raamatute kohta
Püsilink: https://www.kriso.ee/db/97813510497336e.html
Märksõnad: