Principal polynomial analysis
- PMID: 25164247
- DOI: 10.1142/S0129065714400073
Principal polynomial analysis
Abstract
This paper presents a new framework for manifold learning based on a sequence of principal polynomials that capture the possibly nonlinear nature of the data. The proposed Principal Polynomial Analysis (PPA) generalizes PCA by modeling the directions of maximal variance by means of curves, instead of straight lines. Contrarily to previous approaches, PPA reduces to performing simple univariate regressions, which makes it computationally feasible and robust. Moreover, PPA shows a number of interesting analytical properties. First, PPA is a volume-preserving map, which in turn guarantees the existence of the inverse. Second, such an inverse can be obtained in closed form. Invertibility is an important advantage over other learning methods, because it permits to understand the identified features in the input domain where the data has physical meaning. Moreover, it allows to evaluate the performance of dimensionality reduction in sensible (input-domain) units. Volume preservation also allows an easy computation of information theoretic quantities, such as the reduction in multi-information after the transform. Third, the analytical nature of PPA leads to a clear geometrical interpretation of the manifold: it allows the computation of Frenet-Serret frames (local features) and of generalized curvatures at any point of the space. And fourth, the analytical Jacobian allows the computation of the metric induced by the data, thus generalizing the Mahalanobis distance. These properties are demonstrated theoretically and illustrated experimentally. The performance of PPA is evaluated in dimensionality and redundancy reduction, in both synthetic and real datasets from the UCI repository.
Keywords: Principal Polynomial Analysis; classification; coding; dimensionality reduction; manifold learning.
Similar articles
-
Principal components analysis competitive learning.Neural Comput. 2004 Nov;16(11):2459-81. doi: 10.1162/0899766041941880. Neural Comput. 2004. PMID: 15476607
-
Dynamic competitive probabilistic principal components analysis.Int J Neural Syst. 2009 Apr;19(2):91-103. doi: 10.1142/S0129065709001860. Int J Neural Syst. 2009. PMID: 19496205
-
Probabilistic PCA self-organizing maps.IEEE Trans Neural Netw. 2009 Sep;20(9):1474-89. doi: 10.1109/TNN.2009.2025888. Epub 2009 Aug 18. IEEE Trans Neural Netw. 2009. PMID: 19695998
-
Nonlinear complex-valued extensions of Hebbian learning: an essay.Neural Comput. 2005 Apr;17(4):779-838. doi: 10.1162/0899766053429381. Neural Comput. 2005. PMID: 15829090 Review.
-
Linear, nonlinear or categorical: how to treat complex associations in regression analyses? Polynomial transformations and fractional polynomials.Int J Public Health. 2013 Feb;58(1):157-60. doi: 10.1007/s00038-012-0362-0. Epub 2012 May 9. Int J Public Health. 2013. PMID: 22569914 Review. No abstract available.
Cited by
-
Sequential Learning of Principal Curves: Summarizing Data Streams on the Fly.Entropy (Basel). 2021 Nov 18;23(11):1534. doi: 10.3390/e23111534. Entropy (Basel). 2021. PMID: 34828234 Free PMC article.
-
Kernel methods and their derivatives: Concept and perspectives for the earth system sciences.PLoS One. 2020 Oct 29;15(10):e0235885. doi: 10.1371/journal.pone.0235885. eCollection 2020. PLoS One. 2020. PMID: 33119617 Free PMC article.
-
Visual aftereffects and sensory nonlinearities from a single statistical framework.Front Hum Neurosci. 2015 Oct 13;9:557. doi: 10.3389/fnhum.2015.00557. eCollection 2015. Front Hum Neurosci. 2015. PMID: 26528165 Free PMC article.
Publication types
MeSH terms
LinkOut - more resources
Full Text Sources
Other Literature Sources