Hierarchical algorithms

There is a wide variety of hierarchical algorithms available and it is impossible to discuss all of them here. Therefore, we shall only explain the most typical ones, namely the single linkage, the complete linkage and the average linkage methods. 

Let us now cluster the objects of Table 30.8 with a non-hierarchical algorithm. Instead of clustering by joining objects successively, one wants to determine... 

J. R. Gunn, A. Monge, R. A. Friesner, and C. H. Marshall, /. Phys. Chem., 98,702 (1994). Hierarchical Algorithm for Computer Modeling of Protein Tertiary Structure Folding of Myoglobin to 6.2 A Resolution. 

Two standard methods are in common use in the MD community the reaction field method [79,80] and the Ewald summation technique [72,81-83]. There are also various hierarchical algorithms which are quite attractive in principle, but have proved to be difficult to implement efficiently in practice [67,84-87]. An alternative and potentially development interesting complement, is the summation formula developed by Lekner [88,89] which has been given an alternative and more general derivation by Sperb [90]. 

Hierarchical algorithms are characterized by the construction of a tree-like structure. A single linkage algorithm starts with placing in the first cluster the two nearest points. At the next stage, a third point joins the already formed cluster of two if its shortest distance to the members of the first cluster is smaller than to any other point. 

Other hierarchical algorithms are also known, such as that using the average linkage option. In this algorithm, the first set of clusters is done in the same way as previously described. However, the addition of the third element to a cluster considers the average distance to the elements in the previously formed cluster, and the distance between clusters is the average distance from points in the first cluster and the second cluster. 

The hierarchical clustering techniques have been applied to many studies of gene expression patterns with some success.57 However, the hierarchical tree cannot determine the optimal number of clusters in the dataset. The limitation of the hierarchical algorithms is that the number of classes is determined by cutting the tree structure at an ad hoc level selected by the user. Such an ad hoc level does not necessarily reflect the true nature of the underlying structure of the gene expression data. 

Hierarchical Algorithm for Computer Modeling of Protein Tertiary Structure Folding of Myoglobin to 6.2 A Resolution. 

In this paper, we present a systematic approach for simultaneous optimisation of product portfolio and multi-site capacity planning in the face of clinical trials uncertainty while considering the trading structure of the company. A hierarchical algorithm is also proposed for the solution of the resulting large-scale MTT.P model. 

Another problem with such algorithms is that of determining the optimal number of classes that correspond to the cluster substructure of the data set. There are two approaches The use of validity functionals, which is a postfactum method, and the use of hierarchical algorithms, which produce not only the optimal number of classes (based on the needed granularity), but also a binary hierarchy that shows the existing relationships between the classes. 

A trivial application of this principle, however, is the use in the hierarchical algorithm of a list of allowed pairs in generating new segments. This eliminates the need for a scoring function to penalize unfavorable regions of the Ramachandran map, as well as the need to sample such unlikely regions of the conformational space. Although the definition of this list is entirely empirical, based on observation of the PDB, it still represents real interactions that a new structure would be very unlikely to violate. 

Figure 9.20 Hierarchical algorithm for computing the linear relaxation of the arbitrary comb-branched polymer illustrated in (a). The molecule consists of arms and backbone segments. Initially, after a step strain, only the arms can move, by primitive path fluctuationsy from the arm tips Inward. When an arm fully relaxes, it is pruned away, and replaced by a bead at the branch point to represent the frictional drag contributed by that arm see (b). Continued arm relaxation converts the molecule into a star (c), and finally a linear chain (d).The linear chain can complete its relaxation by reptation. From Park and Larson [49].

Park, S.-J., Larson, R. G. A hierarchical algorithm for predicting the linear viscoelastic Properties of polymer melts with long-chain branching. Rheol. Acta (2004) 44, pp. 319-330... 

---