On the use of a Modified Latin Hypercube Sampling (MLHS) method in the estimation of a Mixed Logit Model for vehicle choice

By Stephane Hess, Kenneth Train, John W. Polak. 2006

501 citations recorded by the source corpus.

Subfields: Automotive Engineering, Economics and Econometrics, Transportation

View Paper: https://www.semanticscholar.org/paper/64c1fafa812cf0667d4c715ae01ebbefc454f1be

How researchers see this paper

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

In Automotive Engineering

In Economics and Econometrics

In Transportation

Citations in the corpus

Valuing Water Quality as a Function of Water Quality Measures (2009, unknown): “12Modified Latin Hypercube Sampling (MLHS) is used in the simulation process (Hess, Train, and Polak 2006).”
Willingness-to-pay for renewable energy: Primary and discretionary choice of British households' for micro-generation technologies (2010, methodology): “For the purpose of estimation marginal probability integrals were all approximated numerically by means of simulation methods (Train 2003) based on Modified Hypercube Sampling draws (Hess et al. 2006).”
It’s not that I don’t care, I just don’t care very much: confounding between attribute non-attendance and taste heterogeneity (2013, methodology): “The estimation made use of 250 MLHS draws (Hess et al., 2006) per respondent and per random parameter, where Lognormal distributions were used.”
Discrete Choice Experiments: A Guide to Model Specification, Estimation and Software (2017, methodology): “10 In models with several random coefficients, alternative approaches such as shuffled or scrambled Halton draws [50] or Sobol draws [51, 52] are sometimes used to minimize the correlation between the draws, which can be substantial for standard Halton draws in higher dimensions.”
Reliable estimation of random coefficient logit demand models (2017, methodology): “In contrast, with our most accurate integration approach, which uses 10,000 modified latin hypercube sampling draws (MLHS draws, Hess et al. (2006)) we obtain convergence to a local minimum for 100 percent of the estimations and a coe cient of variation of less than 3 percent.”
Approximation of bayesian efficiency in experimental choice designs (2008, unknown): “The MLHS method (Hess et al. 2005) produces multi-dimensional sequences by combining randomly shuffled versions of one-dimensional sequences made up of 4 The theoretical lowest rate of convergence for quasi-random MC simulation is O (( logK R ) /R ) which depends on the number of dimensions, K,…”
Benefit Estimates for Landscape Improvements: Sequential Bayesian Design and Respondents’ Rationality in a Choice Experiment (2005, methodology): “Such probability is approximated in estimation by simulating the log-likelihood with Halton and shuffled Halton draws (Train 2003; Hess, Train, and Polack 2006 ).”
Assessing the effect of autonomous driving on value of travel time savings: A comparison between current and future preferences (2019, methodology): “The distribution of the random parameters was simulated using 8,000 MLHS (Modified Latin Hypercube Sampling) draws (Hess et al., 2006); the large number of draws was decided on to reflect the high number of dimensions over which the likelihood function must be integrated.”
Estimation on stated-preference experiments constructed from revealed-preference choices (2008, methodology): “…e are constructed analytically from draws from a uniform (as opposed to by accept–reject methods), variance reduction procedures can readily be applied, such as Halton draws (Bhat, 2001; Train, 2003), (t,m,s)-nets (Sándor and Train, 2004), and modified Latin hypercube sampling (Hess et al., 2006).”
Asymmetric preference formation in willingness to pay estimates in discrete choice models (2008, methodology): “All models were coded and estimated in Ox 4.2 (Doornik, 2001), and the errorcomponents models were estimated using 2,000 modified latin hypercube sampling (MLHS) draws (Hess et al., 2006) 12 .”
Best practices for differentiated products demand estimation with PyBLP (2020, methodology): “For example, Hess et al. (2006) and Brunner et al. (2017) employ Modified Latin Hypercube Sampling (MLHS) for MSL and BLP problems, respectively.”
Latent class structures: taste heterogeneity and beyond (2014, background): “Improvements in computer performance as well as the way in which draws from f (β | Ω) can be generated to better represent the distribution (see e.g. Bhat, 2001, 2003; Hess et al., 2006) have led to widespread use of the model in many fields.”