Corpus ID: 14008705

Analysis of Kernel Mean Matching under Covariate Shift

@article{Yu2012AnalysisOK,
  title={Analysis of Kernel Mean Matching under Covariate Shift},
  author={Y. Yu and Csaba Szepesvari},
  journal={ArXiv},
  year={2012},
  volume={abs/1206.4650}
}
  • Y. Yu, Csaba Szepesvari
  • Published 2012
  • Computer Science, Mathematics
  • ArXiv
  • In real supervised learning scenarios, it is not uncommon that the training and test sample follow different probability distributions, thus rendering the necessity to correct the sampling bias. Focusing on a particular covariate shift problem, we derive high probability confidence bounds for the kernel mean matching (KMM) estimator, whose convergence rate turns out to depend on some regularity measure of the regression function and also on some capacity measure of the kernel. By comparing KMM… CONTINUE READING
    63 Citations
    Robust Importance Weighting for Covariate Shift
    • PDF
    Kernel Robust Bias-Aware Prediction under Covariate Shift
    • PDF
    Efficient Sampling-Based Kernel Mean Matching
    • 5
    • PDF
    Doubly Robust Covariate Shift Correction
    • 27
    • PDF
    Robust Classification Under Sample Selection Bias
    • 79
    • PDF
    Covariate Shift in Hilbert Space: A Solution via Sorrogate Kernels
    • 57
    • PDF
    Robust Covariate Shift Regression
    • 32
    • PDF
    Maximum Mean Discrepancy for Class Ratio Estimation: Convergence Bounds and Kernel Selection
    • 48
    • Highly Influenced
    • PDF
    Discriminative Density-ratio Estimation
    • 3
    • PDF
    Covariate Shift: A Review and Analysis on Classifiers
    • 1

    References

    SHOWING 1-10 OF 25 REFERENCES
    Discriminative Learning Under Covariate Shift
    • 237
    • PDF
    Direct Importance Estimation with Model Selection and Its Application to Covariate Shift Adaptation
    • 623
    • Highly Influential
    • PDF
    Learning and evaluating classifiers under sample selection bias
    • 630
    • PDF
    Correcting Sample Selection Bias by Unlabeled Data
    • 1,203
    • Highly Influential
    • PDF
    A Least-squares Approach to Direct Importance Estimation
    • 347
    • PDF
    Learning Bounds for Domain Adaptation
    • 336
    • PDF
    Sample Selection Bias Correction Theory
    • 225
    • Highly Influential
    • PDF
    Statistical analysis of kernel-based least-squares density-ratio estimation
    • 67
    • Highly Influential
    • PDF
    Learning Bounds for Importance Weighting
    • 183
    • PDF