Dataset Shift in Machine Learning

Dataset shift is a common problem in predictive modeling that occurs when the joint distribution of inputs and outputs differs between training and test stages. Covariate shift, a particular case of dataset shift, occurs when only the input distribution changes. Dataset shift is present in most prac...

Celý popis

Uložené v:
Podrobná bibliografia
Hlavní autori: Quinonero-Candela, Joaquin, Sugiyama, Masashi, Schwaighofer, Anton, Lawrence, Neil D
Médium: E-kniha Kniha
Jazyk:English
Vydavateľské údaje: Cambridge, Mass MIT Press 2008
The MIT Press
Vydanie:1
Edícia:Neural Information Processing series
Predmet:
Computer Science
Computing and Processing
Machine learning
Machine Learning & Neural Networks
Mathematical models
ISBN:0262170051, 9780262170055
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Obsah:
  • Intro -- Contents -- Series Foreword -- Preface -- I - Introduction to Dataset Shift -- 1 - When Training and Test Sets Are Di erent: Characterizing Learning Transfer -- 2 - Projection and Projectability -- II - Theoretical Views on Dataset and Covariate Shift -- 3 - Binary Classi cation under Sample Selection Bias -- 4 - On Bayesian Transduction: Implications for the Covariate Shift Problem -- 5 - On the Training/Test Distributions Gap: A Data Representation Learning Framework -- III - Algorithms for Covariate Shift -- 6 - Geometry of Covariate Shift with Applications to Active Learning -- 7 - A Conditional Expectation Approach to Model Selection and Active Learning under Covariate Shift -- 8 - Covariate Shift by Kernel Mean Matching -- 9 - Discriminative Learning under Covariate Shift with a Single Optimization Problem -- 10 - An Adversarial View of Covariate Shift and a Minimax Approach -- IV - Discussion -- 11 - Author Comments -- References -- Notation and Symbols -- Contributors -- Index