Optimization by Simulated Annealing

By S. Kirkpatrick, C. D. Gelatt, M. Vecchi. 1983

How researchers see this paper

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

In Optimization & Training

Citations in the corpus

Generative Modeling by Estimating Gradients of the Data Distribution (2019, methodology): “1) to generate samples, inspired by simulated annealing [26] and annealed importance sampling [32].”
Annealed importance sampling (1998, methodology): “The method of simulated annealing was introduced by Kirkpatrick, Gelatt, and Vecchi (1983) as a way of handling multiple modes in an optimization context.”
Automated Machine Learning: State-of-The-Art and Open Challenges (2019, methodology): “Simulated Annealing is a hyper-parameter optimization approach which is inspired by the metallurgy technique of heating and controlled cooling of materials [61].”
Auto-Keras: An Efficient Neural Architecture Search System (2018, methodology): “Inspired by various heuristic search algorithms for exploring the tree-structured search space and optimization methods balancing between exploration and exploitation, a new method based on A* search [19] and simulated annealing [23] is proposed.”
Neural Combinatorial Optimization with Reinforcement Learning (2016, methodology): “ge set of heuristics, including 2-opt (Johnson, 1990) and Lin–Kernighan (S. Lin, 1973), stopping when it reaches a local minimum. Meta-heuristics combine local search with either simulated annealing (Kirkpatrick et al., 1983), tabu search (Glover & Laguna, 2013) or guided local search (Voudouris”
Adding Gradient Noise Improves Learning for Very Deep Networks (2015, background): “In this sense, we suspect adding gradient noise is similar to simulated annealing (Kirkpatrick et al., 1983) which exploits random noise to explore complex optimization landscapes.”
Modular meta-learning (2018, methodology): “We use simulated annealing [23] to perform this search: it is a local search strategy that employs stochasticity to avoid local optima and has asymptotic optimality guarantees.”
Incorporating Non-local Information into Information Extraction Systems by Gibbs Sampling (2005, methodology): “We test the effectiveness of our technique on two established datasets: the CoNLL 2003 English named entity recognition dataset, and the CMU Seminar Announcements information extraction dataset.”
Protein structure prediction using multiple deep neural networks in the 13th Critical Assessment of Protein Structure Prediction (CASP13) (2019, neutral): “Twooutofthethreemethods(AandB)usedasimulatedannealing 17 based search method.”
Deep Learning of Representations: Looking Forward (2013, methodology): “The idea of tempering (Iba, 2001) for MCMCs is analogous to the idea of simulated annealing (Kirkpatrick et al., 1983) for optimization, and it is designed for and looks very appealing to solve the mode mixing problem: consider a smooth version (higher temperature, obtained by just dividing the…”
Community detection in graphs (2009, methodology): “The minimization of H is carried out via simulated annealing ((Kirkpatrick et al., 1983) and Section VI.A.2), starting from a configuration where spins are randomly assigned to the vertices and the number of states q is very high.”
Programs as Black-Box Explanations (2016, methodology): “The combinatorial optimization is solved using simulated annealing [9] with a logarithmically decreasing temperature schedule, and a proposal function that randomly grows, shrinks, or replaces nodes in the express tree to create valid perturbed expression trees.”
Solving Linear Inverse Problems Using the Prior Implicit in a Denoiser (2020, background): “Since the family of observation densities, pσ(y) forms a scale-space representation of p(x), the algorithm may be viewed as an adaptive form of coarse-to-fine optimization [35–38].”
Logic-Guided Data Augmentation and Regularization for Consistent Question Answering (2020, neutral): “We perform annealing (Kirkpatrick et al., 1983; Li et al., 2019; Du et al., 2019).”
Fusing points and lines for high performance tracking (2005, methodology): “Generalized Monte-Carlo based techniques such as simulated annealing[7] escape local optima easier by randomly”