Regularized Pre-image Estimation for Kernel PCA De-noising
- 289 Downloads
The main challenge in de-noising by kernel Principal Component Analysis (PCA) is the mapping of de-noised feature space points back into input space, also referred to as “the pre-image problem”. Since the feature space mapping is typically not bijective, pre-image estimation is inherently illposed. As a consequence the most widely used estimation schemes lack stability. A common way to stabilize such estimates is by augmenting the cost function by a suitable constraint on the solution values. For de-noising applications we here propose Tikhonov input space distance regularization as a stabilizer for pre-image estimation, or sparse reconstruction by Lasso regularization in cases where the main objective is to improve the visual simplicity. We perform extensive experiments on the USPS digit modeling problem to evaluate the stability of three widely used pre-image estimators. We show that the previous methods lack stability in the is non-linear regime, however, by applying our proposed input space distance regularizer the estimates are stabilized with a limited sacrifice in terms of de-noising efficiency. Furthermore, we show how sparse reconstruction can lead to improved visual quality of the estimated pre-image.
KeywordsKernel PCA Pre-image Regularization De-noising Sparsity
This work is supported in part by the Lundbeckfonden through the Center for Integrated Molecular Brain Imaging (Cimbi), www.cimbi.org.
- 1.Arias, P., Randall, G., & Sapiro, G. (2007). Connecting the out-of-sample and pre-image problems in kernel methods. In IEEE computer society conference on computer vision and pattern recognition (pp. 1–8). Minneapolis, MN.Google Scholar
- 2.Bakir, G. H., Weston, J., & Schölkopf, B. (2004). Learning to find pre-images. In S. L. Thrun, & B. Schölkopf (Eds.), Advances in neural information processing systems (Vol. 16, pp. 449–456). Cambridge: MIT.Google Scholar
- 4.Dambreville, S., Rathi, Y., & Tannenbaum, A. (2006). Statistical shape analysis using kernel PCA. In IS&T/SPIE symposium on electrical imaging.Google Scholar
- 5.Friedman, J. H. (2008). Fast sparse regression and classification. Tech. rep., Department of Statistics, Stanford University.Google Scholar
- 9.Mika, S., Schölkopf, B., Smola, A., Müller, K. R., Scholz, M., & Rätsch, G. (1999). Kernel PCA and de-noising in feature spaces. In M. S. Kearns, S. A. Solla, & D. A. Cohn (Eds.), Advances in neural information processing systems (Vol. 11, pp. 536–542). Cambridge: MIT.Google Scholar
- 10.Nguyen, M. H., & la Torre, F. D. (2009). Robust kernel principal component analysis. Advances in Neural Information Processing Systems, 21, 1185–1192.Google Scholar
- 14.Schölkopf, B., Smola, A. J., Knirsch, P., & Burges, C. J. C. (1998). Fast approximation of support vector kernel expansions, and an interpretation of clustering as approximation in feature spaces. In P. Levi, M. Schanz, R.-J. Ahlers, & F. May (Eds.), Mustererkennung 1998, 20. DAGM-Symposium (pp. 125–132). Berlin: Springer.Google Scholar
- 15.Takahashi,T., & Kurita, T. (2002). Robust de-noising by kernel PCA. In ICANN ’02: Proceedings of the international conference on artificial neural networks (pp. 739–744). Berlin: Springer.Google Scholar
- 17.Thorstensen, N., Segonne, F., & Keriven, R. (2008). Normalization and preimage problem in gaussian kernel PCA. In 2008 15th IEEE international conference on image processing—ICIP 2008, (pp. 741–744). San Diego, California.Google Scholar
- 19.Tikhonov, A. N., & Arsenin, V. Y. (1977). Solution of ill-posed problems. New York: Wiley.Google Scholar
- 20.Zheng, W. S., & Lai, J. H. (2006). Regularized locality preserving learning of pre-image problem in kernel principal component analysis. International Conference on Pattern Recognition, 2, 456–459.Google Scholar