Complex Systems
Current Issue

Volume 26, Number 3 (2017)


A New Kind of Science: A 15-Year View
Stephen Wolfram

Starting now, in celebration of its 15th anniversary, A New Kind of Science will be freely available in its entirety, with high-resolution images, on the web or for download.


Reservoir Computing Using Nonuniform Binary Cellular Automata
Stefano Nichele and Magnus S. Gundersen

The reservoir computing (RC) paradigm utilizes a dynamical system (a reservoir) and a linear classifier (a readout layer) to process data from sequential classification tasks. In this paper, the usage of cellular automata (CAs) as a reservoir is investigated. The use of CAs in RC has been showing promising results. In this paper, it is shown that some cellular automaton (CA) rules perform better than others and the reservoir performance is improved by increasing the size of the CA reservoir itself. In addition, the usage of parallel loosely coupled (nonuniform) CA reservoirs, where each reservoir has a different CA rule, is investigated. The experiments performed on nonuniform CA reservoirs provide valuable insights into CA reservoir design. The results herein show that some rules do not work well together, while other combinations work remarkably well. This suggests that nonuniform CAs could represent a powerful tool for novel CA reservoir implementations.


The "Two's Company, Three's a Crowd" Game
Philippe Collard

The so-called Two's Company, Three's a Crowd game is a tiny artificial world populated by individuals, each with their own behavior, which is expressed by the way they move around the world; when they move, individuals meet others and establish social links with some of them. This model allows carrying out experiments in silico; its goal is not truly to model the real world but rather to suggest that a system of individuals, moving through an artificial world and reacting together, is adequate to account for the formation of some patterns comparable to those resulting from animal and human behavior. First, according to the density of individuals and the distribution of mobility behaviors, we study the properties of the resulting relational network. Then, assuming that in turn, proximity links may influence the behaviors, we study the impact of the feedback loop on both spatial distribution and social patterns. Such dynamics lead to various kinds of homophilous groups where links between separate groups are weak, while links within a group are strong. Although the emergent social networks could be seen as the result of individual strategies striving for uniformity, seclusion, gregarious instinct or the need to live as a couple or in a narrow group, it is suggested that the explanation does not require a reductionist theory.


Behavior Classification for Turing Machines
Nestor Diaz

A classification for Turing machines is described. Quantitative descriptors for Turing machine behavior are used for measuring repetitiveness, periodicity, complexity and entropy. These descriptors allowed identifying several kinds of behavior for Turing machines, using an approach based on machine learning. The classification was tested and generality was probed over different configurations of Turing machines.

Current issue Subscribe

Complex Systems ISSN 0891-2513
© 1987–2017
Complex Systems Publication, Inc.

Published four times annually
Complex Systems Publications, Inc.
P.O. Box 6149
Champaign, IL 61826 USA

info@complex-systems.com

Subscribe Now

Contribute

Join the leading edge of complex systems research today. There are no publication charges. Authors are provided with 25 reprints of papers published in Complex Systems. Papers may be submitted via the web or email. Find out more »