Segmentation Classifier methods

Region segmentation techniques can be classified into four methods split and merge... 

The linkage of conventional low molar mass Lc s to a linear polymer main chain via a flexible spacer provides a method to realize systematically the liquid crystalline state in linear polymers. Above the glass transition temperature Tg the polymer main chain can be assumed to exhibit, at least in the nematic state, an almost free motion of the chain segments, causing a tendency towards a statistical chain conformation. Due to their mobility, the polymer main chains are able to diffuse past each other, which is a condition to obtain the liquid state. Therefore such polymers can be classified as liquids of high viscosity10O). 

Most time-domain methods are based on this simple idea, and differ only in the choice of the segment durations, splicing times and weighting windows. They can be classified according to whether they make use of pitch information or not. 

Although supervised and unsupervised methods have been in the past well-separated families, a number of hybrid approaches have appeared that combine both unsupervised and supervised classifiers for image segmentation purposes [61]. 

In recent years, as the size of the PDB has increased, database methods have continued to attract attention. With a larger database, recurring structural motifs have been classified for loop structures [111, 114, 116, 129, 134], including their sequence dependence. Database methods have been applied only for loops of up to 8 residues. Fidelis et al. [126] found that for loops of length greater than 7 there is not likely to be a segment in the PDB that corresponds to the correct loop conformations for the 58 protein they looked at. Some authors report that when the database contains a loop of similar structure to the target to be modeled, their methods perform well, with RMSD values around 1A or better. Otherwise they tend to fail [124, 127, 130]. 

Chain segments move with a definite frequency above the glass transition temperature. Consequently, the frequency of the measurement method used or the deformation time of the sample must determine the numerical value of the glass transition temperature observed. Thus, the methods of measurement are classified as static or dynamic according to the speed of the measurement. 

This method are widely applied in medical imaging as a classifier, in which the weights are determined by using training data and the ANN is then used to segment new data. ANNs can also be used in unsupervised way as a clustering method. Because of the many interconnections used in a neural network, spatial information can be easily incorporated into its classification procedures [7]. 

---