Sentence examples for based on likelihood maximization from inspiring English sources
Our method employs an efficient architecture and a low complexity training algorithm based on likelihood maximization.
The proposed methods use geographical boundary information to construct a graph in which a cluster growing process is performed based on likelihood function maximization.
Estimation of the parameters was based on partial likelihood maximization.
A simpler implementation based on direct likelihood maximization via general-purpose numerical optimization algorithms was also considered and found slightly less powerful.
IsoLasso is based on penalized likelihood maximization like FlipFlop and NSMAP, but starts from a restricted set of isoforms the same set returned by Cufflinks for single-end data.
On the theoretical side, we develop a method for the optimization of the likelihood based on Expectation Maximization (EM) (Dempster et al., 1977).
Haplotypes were estimated using maximum likelihood based on Expectation Maximization algorithm [ 39] at two, three or all loci in the ENW population.
We developed an algorithm that attains maximization of the likelihood based on Expectation Maximization, a well accepted paradigm for the numerical optimization of likehood functions in the presence of unobserved variables.
Given the large heterogeneities in thoracic tissues that can impact CT-based attenuation correction of PET activity concentrations, iterative PET reconstructions based on expectation maximization or maximum likelihood tend to produce fewer streak artifacts than those based on analytic filtered backprojection for lung cancer patient imaging studies.
While in a standard decision tree the observations are split based on the minimization of the entropy of the output variable, in a jump tree the split is performed based on the maximization of the likelihood of the observations x ^ i.
The EM algorithm is a two-step numerical iterative estimation procedure based on maximization of the likelihood function, with the characteristic that for each step the likelihood function does not decrease and that convergence to a stationary point for DGLMs is obtained [ 24].
