In order to deal with the severe case of unlabeled data, a method is proposed based on dynamic time alignment of Gaussian mixture model clusters for matching actions in an unsupervised temporal segmentation.
All images were partitioned into grey and white matter (GM/WM), cerebrospinal fluid (CSF) and background using a modified mixture model cluster analysis, after correcting for non-uniformities in image intensity [20].
To correct for non-uniformity in image intensity, the SPM segmentation employs a mixture model cluster analysis to identify voxel intensities, which matches particular tissue types, in combination with a priori probabilistic knowledge of the spatial distribution of tissues derived from the GM, WM, and CSF of the prior probability images (i.e., the priors) [29], [30].
Finally, using a modified mixture model cluster analysis, normalized images were corrected for non-uniformities in signal intensity and partitioned into grey and white matter, CSF and background [ 6].
The Two stage mixed effects model, when incorporating random slopes, and the Gaussian mixture model by clustering the exposure values, Functional clustering model, and Functional logistic regression model, are different ways of capturing the additional information that repeated measures over time provide that may contribute more than pure exposure measure contributions to preterm birth.
The Two stage mixed effects model, Gaussian mixture model by clustering the exposure values, and Functional clustering model ignore uncertainty in the first step estimation or clustering and thus underestimate the standard error in the final odds ratio, increasing the likelihood of a false positive in the results.
Thus, this method is preferable for the present example compared to Method (Gaussian mixture model by clustering the exposure values).
We therefore developed a mixture model based clustering method for tumor subtype classification that uses hidden Markov model (HMM) to account for the spatial correlation in aCGH data.
