On regression estimation with penalties on the model parameters. I am especially interested when the penalties are sparsifying penalties, and I have more notes to sparse regression.
Here I consider general penalties: ridge etc. At least in principle - I have no active projects using penalties without sparsifying them at the moment.
Why might I use such penalties? One reason would be that \(L_2\) penalties have simple forms for their information criteria, as shown by Konishi and Kitagawa (KoKi08 5.2.4).
See also matrix factorisations, optimisation, multiple testing, concentration inequalities, sparse flavoured icecream.
Ridge penalties, relationship with robust regression etc.
In nonparametric statistics we might estimate simultaneously what look like many, many parameters, which we constrain in some clever fashion, which usually boils down to something we can interpret as a “penalty” on the parameters.
“Penalization” has a genealogy unknown to me, but is probably the least abstruse for common, general usage.
The “regularisation” nomenclature claims descent from Tikhonov, (eg TiGl65 etc) who wanted to solve ill-conditioned integral and differential equations, so it’s somewhat more general. “Smoothing” seems to be common in the spline and kernel estimate communities of Wahba (Wahb90) and Silverman (Silv82) et al, who usually actually want to smooth curves. When you say “smoothing” you usually mean that you can express your predictions as a “linear smoother”/hat matrix, which has certain nice properties in generalised cross validation.
“smoothing” is not a great general term, since penalisation does not necessarily cause “smoothness” - for example, some penalties cause the coefficients to become sparse and therefore, from the perspective of coefficients, it promotes non-smooth vectors.
In every case, you wish to solve an ill-conditioned inverse problem, so you tame it by adding a penalty to solutions you feel one should be reluctant to accept.
What should we regularize to attain specific kinds of solutions?
Here’s one thing i saw recently:
Venkat Chandrasekaran, Learning Semidefinite Regularizers via Matrix Factorization
Abstract: Regularization techniques are widely employed in the solution of inverse problems in data analysis and scientific computing due to their effectiveness in addressing difficulties due to ill-posedness. In their most common manifestation, these methods take the form of penalty functions added to the objective in optimization-based approaches for solving inverse problems. The purpose of the penalty function is to induce a desired structure in the solution, and these functions are specified based on prior domain-specific expertise. We consider the problem of learning suitable regularization functions from data in settings in which prior domain knowledge is not directly available. Previous work under the title of ‘dictionary learning’ or ‘sparse coding’ may be viewed as learning a polyhedral regularizer from data. We describe generalizations of these methods to learn semidefinite regularizers by computing structured factorizations of data matrices. Our algorithmic approach for computing these factorizations combines recent techniques for rank minimization problems along with operator analogs of Sinkhorn scaling. The regularizers obtained using our framework can be employed effectively in semidefinite programming relaxations for solving inverse problems. (Joint work with Yong Sheng Soh)
- Akaike, H. (1973a) Information Theory and an Extension of the Maximum Likelihood Principle. In P. F. Caski (Ed.), Proceeding of the Second International Symposium on Information Theory (pp. 199–213). Budapest: Akademiai Kiado
- Akaike, H. (1973b) Maximum likelihood identification of Gaussian autoregressive moving average models. Biometrika, 60(2), 255–265. DOI.
- Azizyan, M., Krishnamurthy, A., & Singh, A. (2015) Extreme Compressive Sampling for Covariance Estimation. arXiv:1506.00898 [Cs, Math, Stat].
- Banerjee, A., Chen, S., Fazayeli, F., & Sivakumar, V. (2014) Estimation with Norm Regularization. In Z. Ghahramani, M. Welling, C. Cortes, N. D. Lawrence, & K. Q. Weinberger (Eds.), Advances in Neural Information Processing Systems 27 (pp. 1556–1564). Curran Associates, Inc.
- Barron, A. R., Huang, C., Li, J. Q., & Luo, X. (2008) MDL, penalized likelihood, and statistical risk. In Information Theory Workshop, 2008. ITW’08. IEEE (pp. 247–257). IEEE DOI.
- Battiti, R. (1992) First-and second-order methods for learning: between steepest descent and Newton’s method. Neural Computation, 4(2), 141–166. DOI.
- Bühlmann, P., & Geer, S. van de. (2011) Additive models and many smooth univariate functions. In Statistics for High-Dimensional Data (pp. 77–97). Springer Berlin Heidelberg
- Bühlmann, P., & van de Geer, S. (2015) High-dimensional inference in misspecified linear models. arXiv:1503.06426 [Stat], 9(1), 1449–1473. DOI.
- Burman, P., & Nolan, D. (1995) A general Akaike-type criterion for model selection in robust regression. Biometrika, 82(4), 877–886. DOI.
- Candès, E. J., & Fernandez-Granda, C. (2013) Super-Resolution from Noisy Data. Journal of Fourier Analysis and Applications, 19(6), 1229–1254. DOI.
- Candès, E. J., & Plan, Y. (2010) Matrix Completion With Noise. Proceedings of the IEEE, 98(6), 925–936. DOI.
- Cavanaugh, J. E.(1997) Unifying the derivations for the Akaike and corrected Akaike information criteria. Statistics & Probability Letters, 33(2), 201–208. DOI.
- Chen, Y.-C., & Wang, Y.-X. (n.d.) Discussion on “Confidence Intervals and Hypothesis Testing for High-Dimensional Regression”.
- Efron, B. (2004) The Estimation of Prediction Error. Journal of the American Statistical Association, 99(467), 619–632. DOI.
- Flynn, C. J., Hurvich, C. M., & Simonoff, J. S.(2013) Efficiency for Regularization Parameter Selection in Penalized Likelihood Estimation of Misspecified Models. arXiv:1302.2068 [Stat].
- Giryes, R., Sapiro, G., & Bronstein, A. M.(2014) On the Stability of Deep Networks. arXiv:1412.5896 [Cs, Math, Stat].
- Gui, J., & Li, H. (2005) Penalized Cox regression analysis in the high-dimensional and low-sample size settings, with applications to microarray gene expression data. Bioinformatics, 21(13), 3001–3008. DOI.
- Hastie, T. J., & Tibshirani, R. J.(1990) Generalized additive models. (Vol. 43). CRC Press
- Hastie, T. J., Tibshirani, Rob, & Wainwright, M. J.(2015) Statistical Learning with Sparsity: The Lasso and Generalizations. . Boca Raton: Chapman and Hall/CRC
- Hawe, S., Kleinsteuber, M., & Diepold, K. (2013) Analysis operator learning and its application to image reconstruction. IEEE Transactions on Image Processing, 22(6), 2138–2150. DOI.
- Hegde, C., Indyk, P., & Schmidt, L. (2015) A nearly-linear time framework for graph-structured sparsity. In Proceedings of the 32nd International Conference on Machine Learning (ICML-15) (pp. 928–937).
- Hoerl, A. E., & Kennard, R. W.(1970) Ridge Regression: Biased Estimation for Nonorthogonal Problems. Technometrics, 12(1), 55–67. DOI.
- Janson, L., Fithian, W., & Hastie, T. J.(2015) Effective degrees of freedom: a flawed metaphor. Biometrika, 102(2), 479–485. DOI.
- Javanmard, A., & Montanari, A. (2014) Confidence Intervals and Hypothesis Testing for High-dimensional Regression. Journal of Machine Learning Research, 15(1), 2869–2909.
- Kaufman, S., & Rosset, S. (2014) When does more regularization imply fewer degrees of freedom? Sufficient conditions and counterexamples. Biometrika, 101(4), 771–784. DOI.
- Konishi, S., & Kitagawa, G. (1996) Generalised information criteria in model selection. Biometrika, 83(4), 875–890. DOI.
- Konishi, S., & Kitagawa, G. (2008) Information criteria and statistical modeling. . New York: Springer
- Montanari, A. (2012) Graphical models concepts in compressed sensing. Compressed Sensing: Theory and Applications, 394–438.
- Needell, D., & Tropp, J. A.(2008) CoSaMP: Iterative signal recovery from incomplete and inaccurate samples. arXiv:0803.2392 [Cs, Math].
- Rahimi, A., & Recht, B. (2009) Weighted Sums of Random Kitchen Sinks: Replacing minimization with randomization in learning. In Advances in neural information processing systems (pp. 1313–1320). Curran Associates, Inc.
- Shen, X., & Huang, H.-C. (2006) Optimal Model Assessment, Selection, and Combination. Journal of the American Statistical Association, 101(474), 554–568. DOI.
- Shen, X., Huang, H.-C., & Ye, J. (2004) Adaptive Model Selection and Assessment for Exponential Family Distributions. Technometrics, 46(3), 306–317. DOI.
- Shen, X., & Ye, J. (2002) Adaptive Model Selection. Journal of the American Statistical Association, 97(457), 210–221. DOI.
- Silverman, B. W.(1982) On the Estimation of a Probability Density Function by the Maximum Penalized Likelihood Method. The Annals of Statistics, 10(3), 795–810. DOI.
- Simon, N., Friedman, J., Hastie, T., & Tibshirani, R. (2011) Regularization Paths for Cox’s Proportional Hazards Model via Coordinate Descent. Journal of Statistical Software, 39(5).
- Stein, C. M.(1981) Estimation of the Mean of a Multivariate Normal Distribution. The Annals of Statistics, 9(6), 1135–1151. DOI.
- Tikhonov, A. N., & Glasko, V. B.(1965) Use of the regularization method in non-linear problems. USSR Computational Mathematics and Mathematical Physics, 5(3), 93–107. DOI.
- Uematsu, Y. (2015) Penalized Likelihood Estimation in High-Dimensional Time Series Models and its Application. arXiv:1504.06706 [Math, Stat].
- van de Geer, S. (2014) Statistical Theory for High-Dimensional Models. arXiv:1409.8557 [Math, Stat].
- Wahba, G. (1990) Spline Models for Observational Data. . SIAM
- Wu, T. T., & Lange, K. (2008) Coordinate descent algorithms for lasso penalized regression. The Annals of Applied Statistics, 2(1), 224–244. DOI.
- Ye, J. (1998) On Measuring and Correcting the Effects of Data Mining and Model Selection. Journal of the American Statistical Association, 93(441), 120–131. DOI.
- Zhang, C.-H., & Zhang, S. S.(2014) Confidence intervals for low dimensional parameters in high dimensional linear models. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 76(1), 217–242. DOI.
- Zhang, Y., Li, R., & Tsai, C.-L. (2010) Regularization Parameter Selections via Generalized Information Criterion. Journal of the American Statistical Association, 105(489), 312–323. DOI.
- Zou, H., & Hastie, T. (2005) Regularization and variable selection via the elastic net. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 67(2), 301–320. DOI.