Differential Sparse Coding

Prior work has shown that features which appear to be biologically plausible as well as empirically useful can be found by sparse coding with a prior such as a laplacian (L1) that promotes sparsity. We show how smoother priors can preserve the benefits of these sparse priors while adding stability to the Maximum A-Posteriori (MAP) estimate that makes it more useful for prediction problems. Additionally, we show how to calculate the derivative of the MAP estimate efficiently with implicit differentiation. One prior that can be differentiated this way is KL-regularization. We demonstrate its effectiveness on a wide variety of applications, and find that online optimization of the parameters of the KL-regularized model can significantly improve prediction performance.

[1]  Rajat Raina,et al.  Self-taught learning: transfer learning from unlabeled data , 2007, ICML '07.

[2]  J. Tropp.  Algorithms for simultaneous sparse approximation. Part II: Convex relaxation , 2006, Signal Process..

[3]  Tunc Geveci,et al.  Advanced Calculus , 2014, Nature.

[4]  Patrice Y. Simard,et al.  Best practices for convolutional neural networks applied to visual document analysis , 2003, Seventh International Conference on Document Analysis and Recognition, 2003. Proceedings..

[5]  Andrew McCallum,et al.  Using Maximum Entropy for Text Classification , 1999 .

[6]  Bo Pang,et al.  Seeing Stars: Exploiting Class Relationships for Sentiment Categorization with Respect to Rating Scales , 2005, ACL.

[7]  Geoffrey E. Hinton,et al.  Semantic hashing , 2009, Int. J. Approx. Reason..

[8]  Ryan M. Rifkin,et al.  Value Regularization and Fenchel Duality , 2007, J. Mach. Learn. Res..

[9]  Yoshua Bengio,et al.  Greedy Layer-Wise Training of Deep Networks , 2006, NIPS.

[10]  David M. Blei,et al.  Supervised Topic Models , 2007, NIPS.

[11]  Inderjit S. Dhillon,et al.  Clustering with Bregman Divergences , 2005, J. Mach. Learn. Res..

[12]  Guy Le Besnerais,et al.  A new look at entropy for solving linear inverse problems , 1999, IEEE Trans. Inf. Theory.

[13]  Shigeo Abe DrEng.  Pattern Classification , 2001, Springer London.

[14]  Yoshua Bengio,et al.  Gradient-based learning applied to document recognition , 1998, Proc. IEEE.

[15]  Michael S. Lewicki,et al.  A Hierarchical Bayesian Model for Learning Nonlinear Statistical Regularities in Nonstationary Natural Signals , 2005, Neural Computation.

[16]  Gábor Lugosi,et al.  Prediction, learning, and games , 2006 .

[17]  Manfred K. Warmuth,et al.  Exponentiated Gradient Versus Gradient Descent for Linear Predictors , 1997, Inf. Comput..

[18]  Bhiksha Raj,et al.  Sparse Overcomplete Latent Variable Decomposition of Counts Data , 2007, NIPS.

[19]  David J. Field,et al.  Sparse coding with an overcomplete basis set: A strategy employed by V1? , 1997, Vision Research.

[20]  Matthew Brand,et al.  Pattern discovery via entropy minimization , 1999, AISTATS.