A Markov Chain Monte Carlo version of the genetic algorithm Differential Evolution: easy Bayesian computing for real parameter spaces

By C. T. ter Braak. Statistics and computing, 2006

871 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computational Theory and Mathematics

View Paper: https://www.semanticscholar.org/paper/ad87c0dc93695e5c8292a21375d2e4998524db20

How researchers see this paper

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

In Artificial Intelligence

In Computational Theory and Mathematics

Citations in the corpus

Global Climate and Atmospheric Composition of the Ultra-hot Jupiter WASP-103b from HST and Spitzer Phase Curve Observations (2018, methodology): “We used emcee (Foreman-Mackey et al. 2013) to fit the HST/WFC3 light curves and differential evolution Monte Carlo for the Spitzer fits (Braak 2006).”
Accelerating Markov Chain Monte Carlo Simulation by Differential Evolution with Self-Adaptive Randomized Subspace Sampling (2009, methodology): “In pri nci le, this allows using N < d in DREAM, an important advantage over DE-MC that re quires N = 2d chains to be run in parallel [27].”
Inferring probabilistic stellar rotation periods using Gaussian processes (2017, background): “The other two proposals are a ‘differential evolution (DE) MCMC’ proposal (Ter Braak 2006; Nelson, Ford & Payne 2014) and the ‘snooker’ extension of DE (ter Braak & Vrugt 2008).”
Parameter Analysis and Estimates for the MODIS Evapotranspiration Algorithm and Multiscale Verification (2019, methodology): “Alternatively, the differential‐evolution Markov chain (DE‐MC) method (Ter Braak, 2006; Ter Braak & Vrugt, 2008) combines a genetic algorithm (Storn & Price, 1997) with the MCMC method.”
High‐dimensional posterior exploration of hydrologic models using multiple‐try DREAM(ZS) and high‐performance computing (2012, methodology): “…advantages, particularly when dealing with complex posterior distributions involving long tails, correlated parameters, multimo-dality, and numerous local optima [ Gilks et al. , 1994; Liu et al. , 2000; ter Braak , 2006; ter Braak and Vrugt , 2008; Vrugt et al. , 2009; Craiu et al. , 2009].”
Markov chain Monte Carlo simulation using the DREAM software package: Theory, concepts, and MATLAB implementation (2016, methodology): “The proof of this is given in ter Braak and Vrugt (2008). Thus, if pacc(X → Xp) is larger than some uniform label drawn from U(0, 1) then the candidate point is accepted and the ith chain moves to the new position, that is xt = X i p, otherwise xt = xt−1.”
Bio-inspired computation: Where we stand and what's next (2019, background): “Carlo and bio-inspired optimization are also enlightening: the efficiency of complex MCMC simulations has been shown to improve by leveraging DE-based heuristics [432], whereas DE solvers encompassing MCMC elements have been shown to better choose the scale and orientation of the distribution modeli”
Confronting uncertainty in model-based geostatistics using Markov Chain Monte Carlo simulation (2011, methodology): “This scheme is an adaptation of the DE-MC (Differential Evolution Markov Chain, Ter Braak, 2006; Ter Braak and Vrugt, 2008) and SCEM-UA (Shuffled Complex Evolutionary Metropolis) algorithms (Vrugt et al., 2003) and produces a Markov Chain that is ergodic with a unique stationary distribution, the pr”
The GAPS Programme with HARPS-N at TNG - XIV. Investigating giant planet migration history via improved eccentricity and mass determination for 231 transiting planets (2017, positive): “The posterior distributions of the orbital parameters were obtained in a Bayesian framework by means of a di ﬀ erential evolution Markov chain Monte Carlo (DE-MCMC) technique, which is the MCMC version of the di ﬀ erential evolution genetic algo-rithm (ter Braak 2006).”
The nature of the TRAPPIST-1 exoplanets. (2018, methodology): “The evolution of the masses and the autocorrelation functions of 5000 DEMCMC steps with the described sub-step sampling for 100 chains are shown in Figure 5, which shows efficient convergence of the chains.”
PyMC: a modern, and comprehensive probabilistic programming framework in Python (2023, methodology): “Other available inference methods include random-walk Metropolis (including specialized versions for binary and categorical parameters), a slice sampler ( Neal, 2003 ), and Differential Evolution Metropolis sampling ( Ter Braak, 2006 ).”