Abstract
We present a type of spherical neural network (SNN) with bounded sigmoidal activation function and study its interpolation capability. We find that the provided SNN can exactly interpolate the training samples. Furthermore, based on the special structure of the presented SNN, we can bound the interpolation error by the modulus of smoothness of the target function, which is different from the previous results on the spherical scattered data interpolation problem.
Similar content being viewed by others
References
Anastassiou GA (2012) Univariate sigmoidal neural network approximation. J Comput Anal Appl 14:659–690
Barron AR (1993) Universal approximation bounds for superpositions of a sigmoidal function. IEEE Trans Inf Theory 39:930–945
Chen DB (1993) Degree of approximation by superpsitions of a sigmoidal function. Approx Theory Its Appl 9:17–28
Chen ZX, Cao FL (2009) The approximation operators with sigmoidal functions. Comput Math Appl 58:758–765
Costarelli D, Spigler R (2013) Constructive approximation by superposition of sigmidal functions. Anal Theory Appl 29:169–196
Costarelli D, Spigler R (2013) Approximation results for neural network opertors activated by sigmoidal functions. Neural Netw 44:101–106
Costarelli D, Spigler R (2013) Multivariate neural network operators with sigmoidal activation functions. Neural Netw 48:72–77
Cybenko G (1989) Approximation by superpositions of sigmoidal function. Math Control Signals Syst 2:303–314
Ditzian Z (1999) A modulus of smoothness on the unit sphere. J d’Anal Math 79:189–200
Fasshauer GE, Schumaker LL (1998) Scattered data fitting on the sphere. In: Dælen M, Lyche T, Schumaker LL (eds) Mathematical methods for curves and surfaces II. Vanderbilt University Press, Nashville, TN, pp 117–166
Freeden W, Gervens T, Schreiner M (1998) Constructive approximation on the sphere. Oxford University Press, New York
Golitschek M, Light W (2001) Interpolation by polynomials and radial basis functions on spheres. Constr Approx 17:1–18
Hahm N, Hong B (2002) Approximation order to a function in \(\overline{C}(\mathbf{R})\) by superposition of a sigmoidal function. Appl Math Lett 15:591–597
Jetter K, Stökler J, Ward J (1999) Error estimates for scattered data interpolation on spheres. Math Comput 68:733–749
Lascaux P, Theodor T (1986) Analyse Numerique Matricielle Apliquée a l’Art de l’Ingenieur. Masson, Paris
Lenze B (1992) Constructive multivariate approximation with sigmoidal functions and applications to neural networks, Numerical Methods of Approximation Theory. Birkhauser Verlag, Basel
Levesley J, Sun X (2005) Approximation in rough native spaces by shifts of smooth kernels on spheres. J Approx Theory 133:269–283
Lewicki G, Marino G (2003) Approximation by superpositions of a sigmoidal function. J Anal Its Appl 22:463–470
Lin SB, Cao FL, Xu ZB (2011) Essential rate for approximation by spherical neural networks. Neural Netw 24:752–758
Lin SB, Cao FL, Chang XY, Xu ZB (2012) A general radial quasi-interpolation operator on the sphere. J Approx Theory 164:1402–1414
Llanas B, Sainz FJ (2006) Constructive approximate interpolation by nerual networks. J Comput Appl Math 188:283–308
Narcowich FJ, Ward JD (2002) Scattered data interpolation on spheres: error estimates and locally supported basis functions. SIAM J Math Anal 33:1393–1410
Narcowich FJ, Sun XP, Ward JD, Wendland H (2007) Direct and inverse sobolev error estimates forscattered data interpolation via spherical basis functions. Found Comput Math 369–370
Sloan I, Womersley R (2000) Constructive polynomial approximation on the sphere. J Approx Theory 103:91–118
Sun XP, Cheney EW (1997) Fundamental sets of continuous functions on spheres. Constr Approx 13:245–250
Wendland H (2005) Scattered data approximation. In: Cambridge monographs on applied and computational mathematics. Cambridge University Press, Cambridge
Womersley R, Sloan I (2001) How good can polynomial interpolation on the sphere be? Adv Comput Math 14:195–226
Author information
Authors and Affiliations
Corresponding author
Additional information
The research was supported by the National 973 Programming (2013CB329404), and the Key Program of National Natural Science Foundation of China (Grant No. 11131006).
Rights and permissions
About this article
Cite this article
Lin, S., Zeng, J. & Xu, Z. Error Estimate for Spherical Neural Networks Interpolation. Neural Process Lett 42, 369–379 (2015). https://doi.org/10.1007/s11063-014-9361-x
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11063-014-9361-x