Nearest neighbour clustering example

In the first example (46), hydrogen adsorption on different sites of a nickel crystal was studied by considering the interaction of a hydrogen atom with a limited number of metal atoms. The nickel crystal was truncated to obtain the clusters shown in Fig. 31, containing 13,10, 9 and 8 nickel atoms, and representing models for the bulk crystal and for the (111), (100) and (110) surfaces respectively. The nearest-neighbour distance in all clusters was... 

The D20-20% strategy performs as well on the unsorted as well as on the re-sorted landscape with, or without, nearest neighbour mutations. This observation seems to point out that a structural similarity based on simple building block pre-sorting has no, or at least a negligible effect although the biological activities do cluster to a certain extent on the sorted landscape in our example. 

An example of a non-hierarchical clustering method is the Jarvis-Patrick algorithm [Jarvis and Patrick 1973]. The Jarvis-Patrick method uses a nearest-neighbours approach. The nearest neighbours of each conformation are the conformations that are the shortest distance away. Two conformations are considered to be in the same cluster in the Jarvis-Patrick method if they satisfy the following criteria ... 

Scarminio and colleagues (1982) reported a comparison of the use of several pattern recognition methods in the analysis of mineral water samples characterized by the concentration of 18 elements, determined by atomic absorption and emission spectrometry. The result of the application of cluster analysis and SIMCA to this data set is discussed elsewhere (Sections 5.4 and 7.2.2) KNN results are shown in Table 5.4. The performance of KNN in this example is really quite impressive for two regions, the training set samples are completely correctly classified up to five nearest neighbours and the overall success rate is 95 per cent or better... 

For solid mercury the clusters were generated as follows The rhombohedral structure can be viewed as a central atom surrounded by atom shells of various size. There first shell contains 12 atoms, 6 of them at distance dnn = Cl and 6 at dni 1.15a. We select the atoms whieh we want to correlate and surround each of them with atoms of the first shell. If the atoms to be correlated are far-away, we select for the embedding also all atoms found within a cylinder of radius 1.15a around the eoimeetion line. One example of the cluster for the nearest neighbour two-body inerement is shown in Fig. 17. The embedding atoms are deseribed with a LCPP " and only with a s basis set... 

---