Abstract
Most neurons in the central nervous system exhibit all-or-none firing behavior. This makes Boolean Functions (BFs) tractable candidates for representing computations performed by neurons, especially at finer time scales, even though BFs may fail to capture some of the richness of neuronal computations such as temporal dynamics. One biologically plausible way to realize BFs is to compute a weighted sum of products of inputs and pass it through a heaviside step function. This representation is called a Higher Order Neuron (HON). A HON can trivially represent any n-variable BF with \(2^n\) product terms. There have been several algorithms proposed for obtaining representations with fewer product terms. In this work, we propose improvements over previous algorithms for obtaining parsimonious HON representations and present numerical comparisons. In particular, we improve the algorithm proposed by Sezener and Oztop [1] and cut down its time complexity drastically, and develop a novel hybrid algorithm by combining metaheuristic search and the deterministic algorithm of Oztop [2].
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
An earlier version of the paper was posted as arXiv:1504.01167.
- 2.
The vector is constructed as \([1, x_1, x_2, x_2x_1,x_3,x_3x_1,x_3x_2x_1,...,x_{n}...x_2x_1]\). Assignments to (\(x_1, x_2, x_3,...,x_{n}\)) are ordered as (0’s represent 1’s and 1’s represent -1’s): 000...0, 100...0, 010...0, 110...0, ..., 111...1.
- 3.
By inspection, we choose \(k_{max}\) as 5 and further constraint the space to odd-integers.
- 4.
The same equivalence classes and numbering used in [1] is adopted.
References
Sezener, C.E., Oztop, E.: Minimal sign representation of boolean functions: algorithms and exact results for low dimensions. Neural Comput. 27(8) (2015)
Oztop, E.: Sign-representation of boolean functions using a small number of monomials. Neural Netw. 22(7), 938–948 (2009)
Oztop, E.: An upper bound on the minimum number of monomials required to separate dichotomies of \(-\)1, 1\(^{\text{n}}\). Neural Comput. 18(12) (2006)
Amano, K.: New upper bounds on the average PTF density of boolean functions. In: Cheong, O., Chwa, K.-Y., Park, K. (eds.) ISAAC 2010. LNCS, vol. 6506, pp. 304–315. Springer, Heidelberg (2010). doi:10.1007/978-3-642-17517-6_28
Siu, K.Y., Roychowdhury, V., Kailath, T.: Discrete Neural Computation. Prentice Hall, Englewood Cliffs (1995)
Mel, B.W.: Information processing in dendritic trees. Neural Comput. 6 (1994)
Amaldi, E., Kann, V.: On the approximability of finding maximum feasible subsystems of linear systems. In: Enjalbert, P., Mayr, E.W., Wagner, K.W. (eds.) STACS 1994. LNCS, vol. 775, pp. 521–532. Springer, Heidelberg (1994). doi:10.1007/3-540-57785-8_168
Saks, M.E.: Slicing the hypercube. In: London Mathematical Society Lecture Note Series 187: Surveys in Combinatorics. Cambridge University Press (1993)
O’Donnell, R., Servedio, R.: Extremal properties of polynomial threshold functions. In: Eighteenth Annual Conference on Computational Complexity (2003)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this paper
Cite this paper
Sezener, C.E., Oztop, E. (2017). Algorithms for Obtaining Parsimonious Higher Order Neurons. In: Lintas, A., Rovetta, S., Verschure, P., Villa, A. (eds) Artificial Neural Networks and Machine Learning – ICANN 2017. ICANN 2017. Lecture Notes in Computer Science(), vol 10613. Springer, Cham. https://doi.org/10.1007/978-3-319-68600-4_18
Download citation
DOI: https://doi.org/10.1007/978-3-319-68600-4_18
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-68599-1
Online ISBN: 978-3-319-68600-4
eBook Packages: Computer ScienceComputer Science (R0)