Sentence examples for highest marginal likelihood from inspiring English sources

During model selection, we used the harmonic mean of the MCMC chain to determine the model with the highest marginal likelihood [50].

On the other hand, in both subjects the exponential and BSP models fitted the data significantly better than the constant one for the viral sub-population infecting tumor tissues, with the BSP model always displaying the highest marginal likelihood (data not shown).

Since the exponential model was the one with the highest marginal likelihood for both HIV-1A and B Albanian data sets (see supplemental Table S2), we could use the population growth rate, r, which is one of the two free parameters (the other is the number of effective infections Ne) of the model to infer the epidemiological quantity R0.

For each dataset, the best model combination overall (highest marginal likelihood) is indicated, and in each case is (very) strongly preferred to the second best.

Out of the tested models (GTR + Γ, both with strict and relaxed lognormal molecular clocks), the clock model that performed better was the lognormal molecular clock, which yielded the highest marginal likelihood.

Models that allow for seasonal migration rates between SC and NC and the global community (Table  2) and include seasonality based on the destination community (Additional file 2: Methods 2.7) have the highest marginal likelihood (−561.0 to −561.8) compared to non-seasonal models (marginal likelihood = −587.0, −588.4 without and with variable selection, respectively).

Only when the extinction fraction is low (10%), the pure birth model obtains a slightly higher marginal likelihood than the birth-death.

According to these comparisons, the partitioned MrBayes analysis of the concatenated data has a significantly higher marginal likelihood than the unpartitioned analysis of the same data.

A seasonal migration model has a higher marginal likelihood than a non-seasonal one (Table  2), and supports migration that is seasonal both to (BF = 70) and from NA (BF > 150).

For ORF Ash, rejection of the null hypothesis was supported by a high degree of deviation from the strict molecular clock model in the root-to-tip test (CV = 0.16) and a significantly higher marginal likelihood value for the relaxed uncorrelated exponential molecular clock (log10BF >1.3 when compared to strict clock model; Table 4, Table S2).

Bayesian analyses do not always favor more complicated models, because simpler models can actually achieve higher marginal likelihoods [58].