Principled Hybrids of Generative and Discriminative Models

@article{Lasserre2006PrincipledHO,
  title={Principled Hybrids of Generative and Discriminative Models},
  author={Julia A. Lasserre and Christopher M. Bishop and Tom Minka},
  journal={2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06)},
  year={2006},
  volume={1},
  pages={87-94}
}
When labelled training data is plentiful, discriminative techniques are widely used since they give excellent generalization performance. However, for large-scale applications such as object recognition, hand labelling of data is expensive, and there is much interest in semi-supervised techniques based on generative models in which the majority of the training data is unlabelled. Although the generalization performance of generative models can often be improved by ‘training them… CONTINUE READING

Similar Papers

Citations

Publications citing this paper.
SHOWING 1-10 OF 204 CITATIONS

Bottleneck Conditional Density Estimators

Rui Shu Stanford
  • 2016
VIEW 5 EXCERPTS
CITES BACKGROUND & METHODS
HIGHLY INFLUENCED

Initializing Neural Networks Using Restricted Boltzmann Machines

VIEW 4 EXCERPTS
CITES BACKGROUND & METHODS
HIGHLY INFLUENCED

Bottleneck Conditional Density Estimation

VIEW 5 EXCERPTS
CITES BACKGROUND & METHODS
HIGHLY INFLUENCED

A geometric view of conjugate priors

  • Machine Learning
  • 2010
VIEW 9 EXCERPTS
CITES METHODS & BACKGROUND
HIGHLY INFLUENCED

Object classification by fusing SVMs and Gaussian mixtures

VIEW 4 EXCERPTS
CITES BACKGROUND
HIGHLY INFLUENCED

Why Does Unsupervised Pre-training Help Deep Learning?

  • J. Mach. Learn. Res.
  • 2010
VIEW 6 EXCERPTS
CITES BACKGROUND
HIGHLY INFLUENCED

Bias-variance tradeoff in hybrid generative-discriminative models

  • Sixth International Conference on Machine Learning and Applications (ICMLA 2007)
  • 2007
VIEW 10 EXCERPTS
CITES BACKGROUND & METHODS
HIGHLY INFLUENCED

Multi-class Semi-supervised Learning with the e-truncated Multinomial Probit Gaussian Process

  • Gaussian Processes in Practice
  • 2007
VIEW 5 EXCERPTS
CITES BACKGROUND & METHODS
HIGHLY INFLUENCED

FILTER CITATIONS BY YEAR

2006
2019

CITATION STATISTICS

  • 23 Highly Influenced Citations

  • Averaged 12 Citations per year from 2017 through 2019

References

Publications referenced by this paper.
SHOWING 1-10 OF 12 REFERENCES

A discriminative framework for modelling object classes

  • 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)
  • 2005
VIEW 4 EXCERPTS
HIGHLY INFLUENTIAL

Discriminatively trained Markov model for sequence classification

  • Fifth IEEE International Conference on Data Mining (ICDM'05)
  • 2005
VIEW 2 EXCERPTS

Generative versus discriminative methods for object recognition

  • 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)
  • 2005
VIEW 1 EXCERPT

Object categorization by learned universal visual dictionary

  • Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1
  • 2005
VIEW 2 EXCERPTS