Buch, Englisch, 450 Seiten, Format (B × H): 156 mm x 234 mm
Methodology, Applications, and Software
Buch, Englisch, 450 Seiten, Format (B × H): 156 mm x 234 mm
Reihe: Chapman & Hall/CRC Texts in Statistical Science
ISBN: 978-1-4822-4997-2
Verlag: Taylor & Francis
This book covers the methodology, applications, and software used in high-dimensional regression modeling. Data collected in many fields is high-dimensional in the sense that many characteristics, or features, are recorded for each observation. The collection of this kind of data is a relatively recent phenomenon, and it poses many challenges that traditional statistical methods have proven incapable of addressing. During the past decade, penalized regression models have become a widespread and important tool for analyzing these kinds of data sets.
Zielgruppe
This book is intended for researchers and graduate students in statistics, biostatistics, and machine learning.
Autoren/Hrsg.
Weitere Infos & Material
FOUNDATIONS. Introduction. The Lasso. Bias reduction. Stability and ridge-type penalties. INFERENCE. False discovery rates. Confidence intervals and hypothesis tests. Variable selection with FDR control. Resampling approaches to inference. OTHER LIKELIHOOD/LOSS FUNCTIONS. Logistic regression and generalized linear models. Cox regression. Accelerate failure time model. Robust regression. STRUCTURED SPARSITY. Bi-level selection. Fusion penalties. Additive and semiparametric models. Multivariate outcomes. Variable selection for interactions.
