Abstract
A system with many degrees of freedom can be characterized by a covariance matrix; principal components analysis focuses on the eigenvalues of this matrix, hoping to find a lower dimensional description. But when the spectrum is nearly continuous, any distinction between components that we keep and those that we ignore becomes arbitrary; it then is natural to ask what happens as we vary this arbitrary cutoff. We argue that this problem is analogous to the momentum shell renormalization group. Following this analogy, we can define relevant and irrelevant operators, where the role of dimensionality is played by properties of the eigenvalue density. These results also suggest an approach to the analysis of real data. As an example, we study neural activity in the vertebrate retina as it responds to naturalistic movies, and find evidence of behavior controlled by a nontrivial fixed point. Applied to financial data, our analysis separates modes dominated by sampling noise from a smaller but still macroscopic number of modes described by a non-Gaussian distribution.
Similar content being viewed by others
Notes
An alternative formulation treats the smallest eigenvalue separately, as with a mass term in field theory, measuring all eigenvalues by their distance from this minimum. Then \(\rho (\lambda )\) would always have, as \(N\rightarrow \infty \), support near \(\lambda = 0\).
In the analytic discussion of model distributions, above, the natural quantities were the eigenvalues and eigenvectors of the matrix \(K_\mathrm{ij}\). As noted, we don’t have access to this matrix when we are confronted with real data, so we analyze the matrix \(C_\mathrm{ij}\) instead. To emphasize that what we are doing with the data is in the same spirit as the analysis of the models, we abuse notation slightly and recycle the symbols \(\{\lambda _\mu , \, u_\mathrm{i}(\mu )\}\).
In some contexts it would be more natural to look at the distribution of eigenvalues, searching for modes that emerge clearly from a “bulk” that might be ascribed to sampling noise. Plotting eigenvalues vs their rank, as we do here, provides a representation of the cumulative distribution of eigenvalues, and does not require us to make bins along the eigenvalue axis. Rather than plotting from smallest to largest, we plot from largest to smallest, so that the spectra are more directly comparable to a plot of the susceptibility or propagator G(k) vs momentum k in the usual statistical physics examples.
References
Pearson, K.: On lines and planes of closest fit to systems of points in space. Philos. Mag. 2, 559–572 (1901)
Shlens, J.: A tutorial on principal components analysis. arXiv:1404.1100 [cs.LG] (2014)
Hopfield, J.J.: Neural networks and physical systems with emergent collective computational abilities. Proc. Natl. Acad. Sci. USA 79, 2554–2558 (1982)
Wilson, K.G.: Problems in physics with many scales of length. Sci. Am. 241, 158–179 (1979)
Jona-Lasinio, G.: The renormalization group: a probabilistic view. Il Nuovo Cimento 26B, 99–119 (1975)
Wilson, K.G., Kogut, J.: The renormalization group and the \(\epsilon \) expansion. Phys. Rep. 12, 75–200 (1974)
Kadanoff, L.P.: Scaling laws for Ising models near \(T_c\). Physics 2, 263–272 (1966)
Kadanoff, L.P.: From simulation model to public policy: an examination of Forrester’s “Urban Dynamics”. Simulation 16, 261–268 (1971)
Kadanoff, L.P., Weinblatt, H.: Public policy conclusions from urban growth models. IEEE Trans. Syst. Man Cybern. SMC–2, 139–165 (1972)
Bensimon, D., Kadanoff, L.P., Liang, S., Shraiman, B.I., Tang, C.: Viscous flows in two dimensions. Rev. Mod. Phys. 58, 977–999 (1986)
Halsey, T.C., Jensen, M.H., Kadanoff, L.P., Procaccia, I., Shraiman, B.I.: Fractal measures and their singularities: the characterization of strange sets. Phys. Rev. A33, 1141–1151 (1986); erratum 34, 1601 (1986)
Constantin, P., Kadanoff, L.P.: Singularities in complex interfaces. Philos. Trans. R. Soc. Lond. Ser. A 333, 379–389 (1990)
Bertozzi, A., Brenner, M., Dupont, T.F., Kadanoff, L.P.: Singularities and similarities in interface flows. In: Sirovich, L.P. (ed.) Trends and Perspectives in Applied Mathatematics. Springer Verlag Applied Math Series Vol. 100, pp. 155–208 (1994)
Coppersmith, S.N., Blank, R.D., Kadanoff, L.P.: Analysis of a population genetics model with mutation, selection, and pleitropy. J. Stat. Phys. 97, 429–459 (1999)
Povinelli, M.L., Coppersmith, S.N., Kadanoff, L.P., Nagel, S.R., Venkataramani, S.C.: Noise stabilization of self-organized memories. Phys. Rev. E 59, 4970–4982 (1999)
Kadanoff, L.P.: More is the same: mean field theory and phase transitions. J. Stat. Phys. 137, 777–797 (2009)
Kadanoff, L.P.: Relating theories via renormalization. Stud. Hist. Philos. Sci. B 44, 22–39 (2013)
Kadanoff, L.P.: Reflections on Gibbs: from statistical physics to the Amistad. J. Stat. Phys. 156, 1–9 (2014)
Kadanoff, L.P.: Innovations in statistical physics. Annu. Rev. Cond. Matter Phys. 6, 1–14 (2015)
Wilson, K.G., Fisher, M.E.: Critical exponents in 3.99 dimensions. Phys. Rev. Lett. 28, 240–243 (1972)
Amit, D.J., Martin-Mayor, V.: Field Theory, the Renormalization Group, and Critical Phenomena. Graphs to Computers, 3rd edn. World Scientific, Singapore (2005)
Binder, K.: Finite size scaling analysis of Ising model block distribution functions. Z. Phys. B 43, 119–140 (1981)
Tkačik, G., Marre, O., Amodei, D., Schneidman, E., Bialek, W., Berry II, M.J.: Searching for collective behavior in a large network of sensory neurons. PLoS Comput. Biol. 10, e1003408 (2014)
Abarbanel, H.D.I., Brown, R., Sidorowich, J.J., Tsimring, L.S.: The analysis of observed chaotic data in physical systems. Rev. Mod. Phys. 65, 1331–1392 (1993)
Mora, T., Bialek, W.: Are biological systems poised at criticality? J. Stat. Phys. 144, 268–302 (2011)
Tkačik, G., Mora, T., Marre, O., Amodei, D., Palmer, S.E., Berry II, M.J., Bialek, W.: Thermodynamics and signatures of criticality in a network of neurons. Proc. Natl. Acad. Sci. USA 112, 11508–11513 (2015)
Marsili, M.: Dissecting financial markets: sectors and states. Quant. Financ. 2, 297–302 (2002)
Lillo, F., Mantegna, R.N.: Variety and volatility in financial markets. Phys. Rev. E 62, 6126–6134 (2000)
Bouchaud, J.P., Potters, M: Financial applications. In: Akemann, G., Baik, J., Di Francesco, P. (eds.) The Oxford Handbook of Random Matrix Theory. Oxford University Press, Oxford (2011). arXiv:0910.1205 [q–fin.ST] (2009)
Bun, J., Allez, R., Bouchaud, J.P., Potters, M.: Rotational invariant estimator for general noisy matrices. arXiv:1502.06736 [cond–mat.stat–mech] (2015)
Bun, J., Bouchaud, J.-P., Potters, M.: Cleaning large correlation matrices: tools from random matrix theory. arXiv:1610.08104 [cond–mat.stat–mech] (2016)
Aygün, E., Erzan, A.: Spectral renormalization group theory on networks. J. Phys. Conf. Ser. 319, 012007 (2011)
Castellana, M.: Real-space renormalization group analysis of a non-mean-field spin-glass. EPL 95, 47014 (2011)
Angelini, M.C., Parisi, G., Ricci-Tersenghi, F.: Ensemble renormalization group for disordered systems. Phys. Rev. B 87, 134201 (2013)
Angelini, M.C., Biroli, G.: Spin glass in a field: a new zero-temperature fixed point in finite dimensions. Phys. Rev. Lett. 114, 095701 (2015)
Brown, K.S., Hill, C.C., Calero, G.A., Myers, C.R., Lee, K.H., Sethna, J.P., Cerione, R.A.: The statistical mechanics of complex signaling networks: Nerve growth factor aignaling. Phys. Biol. 1, 184–195 (2004)
Waterfall, J.J., Casey, F.P., Gutenkunst, R.N., Brown, K.S., Myers, C.R., Brouwer, P.W., Elser, V., Sethna, J.P.: Sloppy model universality class and the Vandermonde matrix. Phys. Rev. Lett. 97, 150601 (2006)
Gutenkunst, R.N., Waterfall, J.J., Casey, F.P., Brown, K.S., Myers, C.R., Sethna, J.P.: Universally sloppy parameter sensitivities in systems biology. PLoS Comput. Biol. 3, e189 (2007)
Transtrum, M.K., Machta, B.B., Sethna, J.P.: Geometry of nonlinear least squares with applications to sloppy models and optimization. Phys. Rev. E 83, 036701 (2011)
Matcha, B.B., Chachra, R., Transtrum, M.K., Sethna, J.P.: Parameter space compression underlies emergent theories and predictive models. Science 342, 604–607 (2013)
Mehta, P., Schwab, D.J.: An exact mapping between the variational renormalization group and deep learning. arXiv:1410.3831 [stat.ML] (2014)
Acknowledgements
We thank D Amodei, MJ Berry II, and O Marre for making available the data of Ref. [23] and M Marsili for the data of Ref. [27]. We are especially grateful to G Biroli, J–P Bouchaud, MP Brenner, CG Callan, A Cavagna, I Giardina, MO Magnasco, A Nicolis, SE Palmer, G Parisi, and DJ Schwab for helpful discussions and comments on the manuscript. Work at CUNY was supported in part by the Swartz Foundation. Work at Princeton was supported in part by Grants from the National Science Foundation (PHY-1305525, PHY-1451171, and CCF-0939370) and the Simons Foundation.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Bradde, S., Bialek, W. PCA Meets RG. J Stat Phys 167, 462–475 (2017). https://doi.org/10.1007/s10955-017-1770-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10955-017-1770-6