Abstract

AbstractIn Chapter 7 we showed that the saturated-linear activation function is not unique in its Turing universality, but rather that various sigmoidal-like activation functions can form finite-size architectures which are Turing universal as well. The class of activation functions considered in this chapter is much wider than that of the previous chapter, and as a result the lower bound on its computational power is weaker. We prove that any function for which the left and right limits exist and are different can serve as an activation function for the neurons to yield a network that is at least as strong computationally as a finite automaton.

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.