September, 1982 | J. Doyne Farmer, Edward Ott, and James A. Yorke
The paper discusses the dimension of chaotic attractors, focusing on different definitions of dimension and their relevance to understanding chaotic systems. It introduces the concept of "natural measure," which reflects the probability distribution of points on an attractor, and contrasts it with the "fractal dimension," which is based on geometric properties. The authors argue that the natural measure dimension is typically equal to the Lyapunov dimension, which is derived from the stability properties of the system. They also present a generalized Baker's transformation as an example to illustrate these concepts, showing that the natural measure dimension can be calculated more easily than other dimensions. The paper concludes that the natural measure dimension is more important for characterizing chaotic attractors due to its physical relevance and computational feasibility. The study supports the conjecture that the natural measure dimension is equal to the Lyapunov dimension for typical chaotic attractors.The paper discusses the dimension of chaotic attractors, focusing on different definitions of dimension and their relevance to understanding chaotic systems. It introduces the concept of "natural measure," which reflects the probability distribution of points on an attractor, and contrasts it with the "fractal dimension," which is based on geometric properties. The authors argue that the natural measure dimension is typically equal to the Lyapunov dimension, which is derived from the stability properties of the system. They also present a generalized Baker's transformation as an example to illustrate these concepts, showing that the natural measure dimension can be calculated more easily than other dimensions. The paper concludes that the natural measure dimension is more important for characterizing chaotic attractors due to its physical relevance and computational feasibility. The study supports the conjecture that the natural measure dimension is equal to the Lyapunov dimension for typical chaotic attractors.