Euclidean Distance Clustering Research Articles

Comparison of feature selection techniques for ANN-based voltage estimation

Abstract Fast estimation of bus voltage magnitude is essential for security monitoring and analysis of power system. An approach based on a parallel self-organising hierarchical neural network (PSHNN) is proposed to estimate bus voltage magnitudes at all the PQ buses of a power system in an efficient manner. PSHNN is a multi-stage neural network in which stages operate in parallel rather than in series during testing. The revised back-propagation algorithm is used for learning input non-linearities along with forward-backward training of stage neural networks. A method based on Euclidean distance clustering is proposed for feature selection. Effectiveness of the proposed method is compared with two existing methods of feature-selection entropy based and angular distance based clustering methods for bus voltage magnitude estimation at different loading conditions in the IEEE 30-bus system and a practical 75-bus Indian system. The PSHNN based on Euclidean distance based clustering method is found to be superior in terms of training time and error performance.

Multitemporal characterization and mapping of montane sagebrush communities using Indian IRS LISS‐II imagery

The sagebrush landscape of Grand Teton National Park, Wyoming, forms four distinct communities: low sagebrush (Artemisia arbuscula); mountain big sagebrush (Artemisia tridentata); mixed low sagebrush/big sagebrush; and a big sagebrush/bitterbrush (Purshia tridentata) mosaic. Collectively, the sagebrush communities provide a significant source of forage for deer, elk, pronghorn antelope, and bison, particularly in winter. The objective of this research was to characterize seasonal changes in spectral reflectance for four sagebrush communities and to develop a multitemporal mapping procedure for the four sagebrush communities using Indian IRS LISS‐II multispectral satellite imagery. Results indicate that the four sagebrush communities can be best differentiated using visible bands of the IRS‐1B LISS‐II, independent of image date. Near‐infrared reflectance was of value for distinguishing between sites only in late fall (October). Statistically significant differences in NDVI values existed between the four communities only for summer dates (June and August) of imagery. Principal component analysis (PCA) was used to compress eight bands (two dates, four bands per date) of Indian IRS‐IB data into four PCA bands. A euclidean distance clustering algorithm was used to classify the four PCA components into forty spectral clusters. Field data acquired during summer 1996 and from the literature were used to identify and combine spectral clusters to produce a final map of sagebrush community types in Grand Teton National Park. Overall accuracy was 65%, and highest for the mixed big sagebrushAow sagebrush community at 86%.

Serological and biochemical variation among potato strains of Erwinia carotovora subsp. atroseptica and their taxonomic relationship to other E. carotovora strains

The serological and biochemical characteristics of 32 Erwinia carotovora subsp. atroseptica strains from potato were compared with 48 other pectolytic Erwinia strains. Biochemical characteristics were examined by the API 20E and API 50CHE systems. Numerical analysis using the Euclidean distance coefficients and clustering by the unweighted average pair group method indicated that these E. carotovora subsp. atroseptica strains formed a distinct cluster (subphenon A1) that could be differentiated from other E. carotovora strains. Three non‐potato strains also belonged to this group; two of these were from tomato and the other from Chinese cabbage. Named E. carotovora subsp. atroseptica strains from other hosts clustered into other phenons. Sixty‐three per cent of subphenon A1 strains tested in this study typed into serogroup I. One potato strain in another phenon also typed into this serogroup. The subphenon A1 strains that did not type into serogroup I typed into serogroups XVIII, XX, or XXII. Many of these strains, however, expressed several different O antigens which were also expressed by E. carotovora strains in other phenons.

Solar climates of the United States based on long-term monthly averaged daily insolation values

The purpose of this paper is to develop and analyze an objective classification and regionalization scheme of solar climates for the conterminous United States. The use of principal components analysis and Euclidean distance cluster analysis generates 18 solar climatic types based upon long-term mean monthly data. Two of the resultant regions contain 76 per cent of the 221 stations used in this study. The results are shown to vary significantly from the previously published maps of solar climates in the United States.

Twentieth century changes in winter climatic regions

The configuration of the northern hemispheric general atmospheric circulation system shifted from a zonal to a meridional pattern in the early 1950s. Winter climatic regions in the conterminous United States are developed for a ten year period dominated by zonal flow and a second decade of meridional flow using a combination of principal components factor analysis and a Euclidean distance clustering algorithm. The results demonstrate that regional patterns in the surface climatic data substantially changed as the circulation system shifted its basic configuration. The regional structures of the eastern United States and the Great Plains appeared to be particularly sensitive to the change in the upperlevel flow pattern.

STUDIES ON GENETIC DIVERSITY IN Vigna mungo L. Hepper IN YMV HOTSPOT

The present investigation was conducted to examine the 41 blackgram genotypes along with one check (T-9) to study the genetic diversity. Analysis of variance showed highly significant differences among 41blackgram genotypes for 9 quantitative characters studied. Maximum genotypic and phenotypic variance was recorded for percentage of disease infection, single plant seed yield, and number of pods per plant. Minimum GCV and PCV were recorded for pod length, days to 50% flowering, number of seeds per pod. High heritability was recorded for percentage of disease infection, single plant seed yield, and number of pods per plant. High heritability coupled with high genetic advance as percent of mean was recorded for percentage of disease infection, single plant seed yield. Genetic diversity estimated in 41 blackgram genotypes using Mahalanobis’s D2 statistic. Forty one genotypes were grouped into seven clusters by Tocher method (Mahalanobis Euclidean Distance) cluster analysis. The maximum inter-cluster distance was observed between cluster VI and cluster VII and maximum intra-cluster distance was observed in cluster VI. Cluster VII showed maximum cluster mean value for seed yield per plant. Among all the characters, seed yield per plant and percentage of disease infection contributes maximum.