Introduction to Derivative-Free Optimization

By Andrew R. Conn, Katya Scheinberg, Luís Nunes Vicente. 2009

1,008 citations recorded by the source corpus.

Subfields: Geometry and Topology, Numerical Analysis

How researchers see this paper

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

In Geometry and Topology

In Numerical Analysis

Citations in the corpus

Stochastic Convex Optimization with Bandit Feedback (2011, background): “We note that most of the existing approaches to derivative-free optimization outlined in the recent book [7] typically search for a descent or sufficient descent direction and then take a step in this direction.”
Use of reduced-order models in well control optimization (2017, methodology): “Generalized pattern search (GPS) algorithms are described in detail in Audet and Dennis Jr (2002), Kolda et al. (2003), Conn et al. (2009).”
Joint optimization of oil well placement and controls (2012, methodology): “Derivative-free methods [10, 27] have also been shown to perform satisfactorily for the control opti-”
Simulation optimization: a review of algorithms and applications (2014, background): “As explained in [43], some DFO algorithms, under certain assumptions, expect rates that are closer to linear than quadratic, and therefore early termination may be suitable.”
Communication-Efficient Edge AI: Algorithms and Systems (2020, methodology): “ release their internal DNNs structure and weights [148]. In zeroth-order optimization algorithms, the full gradients are typically estimated via gradient estimators based on only the function values [149]. For instance, we can use the quantity (l(w+v) l(w))v=to approximate the gradient of function ”
A Survey of Optimization Methods From a Machine Learning Perspective (2019, neutral): “For the target function L(Θ), whenΘ is already obtained, the jth dimension of Θ is solved by [155] θ j = argminθj∈RL(θ t+1 1 , .”
Nuclear Energy Density Optimization (2010, background): “[76] Andrew R.”
Survey of multifidelity methods in uncertainty propagation, inference, and optimization (2018, methodology): “pplied. In [134], a gradient-free multidelity trust-region framework with radial basis function interpolants as low-delity models is introduced, building on the gradientfree trust-region methods of [47, 48] and the tailored radial basis function modeling of [212]. This leads to an error interpolatio”
Spatially Selective Artificial-Noise Aided Transmit Optimization for MISO Multi-Eves Secrecy Rate Maximization (2013, methodology): “mal solution of (13). In the optimization literature, there are many derivative-free search algorithms for solving one-dimensional optimization problems, e.g., compass or coordinate search (cf. [31], [32]). In practice, we use either uniform sampling or the golden search [33] to obtain a satisfactor”
Learning surrogate models for simulation‐based optimization (2014, methodology): “2 2 2 [2, 2] 2 2 [6, 8] [6, 11] [12, 15] [12, 15] 2 3 3 3 3 3 [5, 12] [5, 10] [12, 14] [12, 14] 2 4 [3, 4] [3, 4] [3, 4] [3, 4] [8, 11] [8, 10] [11, 12] [11, 12] 2 5 [2, 4] [2, 4] [2, 5] [2, 5] [3, 12] [4, 11] [10, 16] [10, 16] 2 6 [5, 6] [6, 6] [5, 6] [6, 6] [7, 10] [6, 7] [11, 13] [11, 13] 2 7 [4,”
Bandit Algorithms (2020, background): “Most well known (and comprehensive) is the book by Cover and Thomas [2012]. Another famous book is the elementary and enjoyable introduction by MacKay [2003]. The approach we have taken for defining and understanding the relative entropy is inspired by an excellent shorter book by Gray [2011].”