Muutke küpsiste eelistusi

Practical Guide to Logistic Regression [Pehme köide]

4.60/5 (8 hinnangut Goodreads-ist)
(California Institute of Technology, Pasadena, and Arizona State University, Tempe, USA)
  • Formaat: Paperback / softback, 174 pages, kõrgus x laius: 216x138 mm, kaal: 254 g, 7 Tables, black and white; 12 Illustrations, black and white
  • Ilmumisaeg: 09-Jul-2015
  • Kirjastus: Chapman & Hall/CRC
  • ISBN-10: 1498709575
  • ISBN-13: 9781498709576
Teised raamatud teemal:
Practical Guide to Logistic RegressionSuurem pilt
  • Formaat: Paperback / softback, 174 pages, kõrgus x laius: 216x138 mm, kaal: 254 g, 7 Tables, black and white; 12 Illustrations, black and white
  • Ilmumisaeg: 09-Jul-2015
  • Kirjastus: Chapman & Hall/CRC
  • ISBN-10: 1498709575
  • ISBN-13: 9781498709576
Teised raamatud teemal:
Püsilink: https://www.kriso.ee/db/9781498709576.html
Märksõnad:
Author Joseph M. Hilbe presents students, academics, researchers, and professionals working in a wide variety of contexts with an examination of the basic logic regression model and its applications in data analysis for a variety of fields. The author has organized the main body of his text in six chapters devoted to statistical models, single predictor logistic models, multiple predicator logistic models, testing and fitting a logistic model, and a wide variety of other related subjects. The author is a faculty member of the California Institute of Technology and Arizona State University. Annotation ©2015 Ringgold, Inc., Portland, OR (protoview.com)

Practical Guide to Logistic Regression covers the key points of the basic logistic regression model and illustrates how to use it properly to model a binary response variable. This powerful methodology can be used to analyze data from various fields, including medical and health outcomes research, business analytics and data science, ecology, fisheries, astronomy, transportation, insurance, economics, recreation, and sports. By harnessing the capabilities of the logistic model, analysts can better understand their data, make appropriate predictions and classifications, and determine the odds of one value of a predictor compared to another.

Drawing on his many years of teaching logistic regression, using logistic-based models in research, and writing about the subject, Professor Hilbe focuses on the most important features of the logistic model. Serving as a guide between the author and readers, the book explains how to construct a logistic model, interpret coefficients and odds ratios, predict probabilities and their standard errors based on the model, and evaluate the model as to its fit. Using a variety of real data examples, mostly from health outcomes, the author offers a basic step-by-step guide to developing and interpreting observation and grouped logistic models as well as penalized and exact logistic regression. He also gives a step-by-step guide to modeling Bayesian logistic regression.

R statistical software is used throughout the book to display the statistical models while SAS and Stata codes for all examples are included at the end of each chapter. The example code can be adapted to readers’ own analyses. All the code is available on the author’s website.

Arvustused

"The book presents many worked examples, and the choice of interesting data sets all of which are available to the reader is one of its greatest assets. Data availability makes it easy for readers to reproduce the examples from the book, and example code is available for R, SAS and Stata: R code is incorporated into the book chapters, and the end of each chapter gives SAS and Stata code." Ulrike Grömping, Beuth University of Applied Sciences Berlin, Journal of Statistical Software, July 2016

" this book is written in an exceptionally clear style An additional selling point of this text is that it introduces new R functions, which can be applied in ones own work, as well as equivalent SAS and Stata code. the emphasis on understanding logistic regression modelling rather than on the mechanistic application of techniques is one of the great strengths of the book. Anyone who reads this book will therefore feel that they have a good understanding of this subject " Significance Magazine, February 2016

"Big Data is ascendant, but even the biggest data often boil down to a decision between two categories: survive or die, purchase or dont purchase, click or dont click, fraudulent or honest, default or pay. Logistic regression is the classic workhorse for this 0/1 data, and Joseph Hilbes new book presents a guide for the practitioner, chock full of useful R, Stata, and SAS code. Hilbe has worked with practitioners and aspiring practitioners in virtually every field that uses statistics, including for over a decade via his courses at Statistics.com. His new book is truly, in his own words, a tutorial between you and me." Peter Bruce, Founder and President of the Institute for Statistics Education at Statistics.com

Preface ix
Author xv
1 Statistical Models
1(12)
1.1 What Is a Statistical Model?
1(2)
1.2 Basics of Logistic Regression Modeling
3(1)
1.3 The Bernoulli Distribution
4(3)
1.4 Methods of Estimation
7(6)
SAS Code
11(1)
Stata Code
12(1)
2 Logistic Models: Single Predictor
13(36)
2.1 Models with a Binary Predictor
13(5)
2.2 Predictions, Probabilities, and Odds Ratios
18(2)
2.3 Basic Model Statistics
20(8)
2.3.1 Standard Errors
20(3)
2.3.2 Z Statistics
23(1)
2.3.3 p-Values
23(1)
2.3.4 Confidence Intervals
24(4)
2.4 Models with a Categorical Predictor
28(4)
2.5 Models with a Continuous Predictor
32(5)
2.5.1 Varieties of Continuous Predictors
32(1)
2.5.2 A Simple GAM
33(1)
2.5.3 Centering
34(2)
2.5.4 Standardization
36(1)
2.6 Prediction
37(12)
2.6.1 Basics of Model Prediction
37(2)
2.6.2 Prediction Confidence Intervals
39(2)
SAS Code
41(6)
Stata Code
47(2)
3 Logistic Models: Multiple Predictors
49(22)
3.1 Selection and Interpretation of Predictors
49(3)
3.2 Statistics in a Logistic Model
52(6)
3.3 Information Criterion Tests
58(3)
3.3.1 Akaike Information Criterion
58(1)
3.3.2 Finite Sample
59(1)
3.3.3 Bayesian Information Criterion
60(1)
3.3.4 Other Information Criterion Tests
60(1)
3.4 The Model Fitting Process: Adjusting Standard Errors
61(4)
3.4.1 Scaling Standard Errors
61(2)
3.4.2 Robust or Sandwich Variance Estimators
63(1)
3.4.3 Bootstrapping
64(1)
3.5 Risk Factors, Confounders, Effect Modifiers, and Interactions
65(6)
SAS Code
67(3)
Stata Code
70(1)
4 Testing and Fitting a Logistic Model
71(36)
4.1 Checking Logistic Model Fit
71(10)
4.1.1 Pearson Chi2 Goodness-of-Fit Test
71(1)
4.1.2 Likelihood Ratio Test
72(1)
4.1.3 Residual Analysis
73(6)
4.1.4 Conditional Effects Plot
79(2)
4.2 Classification Statistics
81(7)
4.2.1 S--S Plot
84(1)
4.2.2 ROC Analysis
84(2)
4.2.3 Confusion Matrix
86(2)
4.3 Hosmer-Lemeshow Statistic
88(3)
4.4 Models with Unbalanced Data and Perfect Prediction
91(2)
4.5 Exact Logistic Regression
93(3)
4.6 Modeling Table Data
96(11)
SAS Code
101(4)
Stata Code
105(2)
5 Grouped Logistic Regression
107(20)
5.1 The Binomial Probability Distribution Function
107(2)
5.2 From Observation to Grouped Data
109(4)
5.3 Identifying and Adjusting for Extra Dispersion
113(2)
5.4 Modeling and Interpretation of Grouped Logistic Regression
115(2)
5.5 Beta-Binomial Regression
117(10)
SAS Code
123(2)
Stata Code
125(2)
6 Bayesian Logistic Regression
127(24)
6.1 A Brief Overview of Bayesian Methodology
127(3)
6.2 Examples: Bayesian Logistic Regression
130(21)
6.2.1 Bayesian Logistic Regression Using R
130(7)
6.2.2 Bayesian Logistic Regression Using JAGS
137(6)
6.2.3 Bayesian Logistic Regression with Informative Priors
143(4)
SAS Code
147(1)
Stata Code
148(1)
Concluding Comments
149(2)
References 151(2)
Index 153
Joseph M. Hilbe is a Solar System Ambassador with NASA's Jet Propulsion Laboratory at the California Institute of Technology, an adjunct professor of statistics at Arizona State University, and an emeritus professor at the University of Hawaii. He also teaches five web-based courses on statistical modeling at Statistics.com. He is president of the International Astrostatistics Association, elected fellow of the American Statistical Association, elected member of the International Statistical Institute, and full member of the American Astronomical Society. Professor Hilbe is one of the world's leading statisticians in modeling discrete and longitudinal data. He has authored 16 books related to statistical modeling, including the best-selling Logistic Regression Models and Modeling Count Data. During the late 1980s and 1990s, Professor Hilbe was a leading figure in the then new area of health outcomes research, serving as director of research at a national chain of hospitals and later CEO of a national health economics firm. He was also on the executive committee forming the Health Policy Statistics Section of the American Statistical Association.