Landscapes and their correlation functions

By Peter F. Stadler. 1996

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

NAS-Bench-101: Towards Reproducible Neural Architecture Search (2019, neutral): “A popular measure of locality is the random-walk autocorrelation (RWA), deﬁned as the autocorrelation of the accuracies of points visited as we perform a long walk of random changes through the space (Weinberger, 1990; Stadler, 1996).”
Local Search is State of the Art for Neural Architecture Search Benchmarks (2020, methodology): “Therefore, we can approximate the PDFs by using the following equation: deﬁned as the autocorrelation of the accuracies of points visited during a random walk on the neighborhood graph [34, 32], and was used to measure locality in NASBench-101 in prior work [39].”
Fitness landscape analysis and memetic algorithms for the quadratic assignment problem (2000, background): “Based on this correlation function, the correlation length [53] of the landscape is defined as”
Local Search is State of the Art for NAS Benchmarks (2020, methodology): “…to the experiment in NASBench-101 (Ying et al., 2019), we compute the random-walk autocorrelation (RWA), a popular measure of locality deﬁned as the autocorrelation of the accuracies of points visited during a walk of random single changes through the search space (Weinberger, 1990; Stadler, 1996).”
Recombination Spaces, Metrics, and Pretopologies (2002, background): “In contrast, point mutation space and the spaces of homologous recombination are weakly metrizable [12, 28 , 29].”
Strategy To Discover Diverse Optimal Molecules in the Small Molecule Universe (2015, unknown): “8,9 To our knowledge, this is the ﬁ rst time that the NKp landscape model was used to compare strategies for chemical space search.”
Metaheuristics in combinatorial optimization: Overview and conceptual comparison (2003, background): “A neighborhood structures the search space in the sense that it defines the topology of the so-called fitness landscape [150, 151, 93, 97], which can be visualized as a labelled graph.”
The evolvability of artificial neural net-works for robot control (2002, background): “A more exact picture, especially when dealing with discrete-valued genotypes, is the connected graph (Stadler, 1996).”
Selection for Protein Stability Enriches for Epistatic Interactions (2018, unknown): “We also note that since L∆Gadd = 430 a∆Glin = a∆Gadd − aβ∅, ∆Gadd is an "elementary" landscape [46,47].”
Adaptation in protein fitness landscapes is facilitated by indirect paths (2016, methodology): “We decomposed the fitness landscape into epistatic interactions of different orders by Fourier analysis 568 [9, 54].”