×

zbMATH — the first resource for mathematics

Automated reverse engineering of nonlinear dynamical systems. (English) Zbl 1155.37044
Summary: Complex nonlinear dynamics arise in many fields of science and engineering, but uncovering the underlying differential equations directly from observations poses a challenging task. The ability to symbolically model complex networked systems is key to understanding them, an open problem in many disciplines. Here we introduce for the first time a method that can automatically generate symbolic equations for a nonlinear coupled dynamical system directly from time series data. This method is applicable to any system that can be described using sets of ordinary nonlinear differential equations, and assumes that the (possibly noisy) time series of all variables are observable. Previous automated symbolic modeling approaches of coupled physical systems produced linear models or required a nonlinear model to be provided manually. The advance presented here is made possible by allowing the method to model each (possibly coupled) variable separately, intelligently perturbing and destabilizing the system to extract its less observable characteristics, and automatically simplifying the equations during modeling. We demonstrate this method on four simulated and two real systems spanning mechanics, ecology, and systems biology. Unlike numerical models, symbolic models have explanatory value, suggesting that automated ‘reverse engineering’ approaches for model-free symbolic nonlinear system identification may play an increasing role in our ability to understand progressively more complex systems in the future.

MSC:
37M10 Time series analysis of dynamical systems
PDF BibTeX XML Cite
Full Text: DOI
References:
[1] Neural networks : the official journal of the International Neural Network Society 2 pp 359– (1989) · Zbl 1383.92015
[2] Neural Computation 3 pp 246– (1991)
[3] J MACHINE LEARNING RES 5 pp 255– (2004)
[4] PROC SPIE 3077 pp 409– (1997)
[5] 6 pp 1341– (1998)
[6] 10 pp 673– (2002)
[7] SYSTEMS SCI 31 pp 5– (2005)
[8] INT J COMPUT COGNIT 3 pp 19– (2005)
[9] KNOWLEDGE INF SYST 8 pp 235– (2005) · Zbl 02224727
[10] King, Nature; Physical Science (London) 427 (6971) pp 247– (2004)
[11] J MACHINE LEARNING RES 6 pp 1651– (2005)
[12] TRANS EVOL COMPUT 9 pp 361– (2005) · Zbl 05451906
[13] Bongard, Science 314 (5802) pp 1118– (2006)
[14] Kirkpatrick, Science 220 (4598) pp 671– (1983) · Zbl 1225.90162
[15] Foster, Nature reviews. Genetics 2 (6) pp 428– (2001)
[16] Milo, Science 303 (5663) pp 1538– (2004)
[17] Strogatz, Nature; Physical Science (London) 410 (6825) pp 268– (2001) · Zbl 1370.90052
[18] MACHINE LEARNING 10 pp 153– (1993) · Zbl 0806.94021
[19] GENETIC PROGRAMMING AND EVOLVABLE MACHINES 3 pp 81– (2002) · Zbl 1052.68579
[20] Weinberger 44 (10) pp 6399– (1991)
[21] 304 pp 178– (2002)
[22] May, Nature; Physical Science (London) 261 (5560) pp 459– (1976) · Zbl 1369.37088
[23] J CONS PERM INT EXP MER 3 pp 3– (1926)
[24] PNAS 97 (22) pp 12079– (2000)
[25] Luscombe, Nature; Physical Science (London) 431 (7006) pp 308– (2004)
[26] Husmeier, Bioinformatics 19 (17) pp 2271– (2003)
[27] J COMP BIOL 7 pp 601– (2000)
[28] Wikswo 153 (4) pp 81– (2006)
[29] Dekel, Nature; Physical Science (London) 436 (7050) pp 588– (2005)
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.