hierarchical clustering

Among different methods for classification, clustering is commonly used methods. Flexible-Beta clustering is successful hierarchical agglomerative clustering which is employed by ecologists as effective clustering method. The aim of the research was to detect the suitable value of beta for flexible-clustering methods. For this purpose, two different forest regions from Hyrcanian and Zagros Oak regions were selected. The clustering algorithms included Flexible-beta algorithms with five value of beta (-0.1, -0.25-, -0.4, -0.6 and -0.8). Five evaluators (Silhouette, MRPP, PARATNA, Phi coefficient) were employed on each cluster solution to evaluate different clustering algorithms. Algorithms were ranked from best to worst on each clustering evaluator for each data set. The results showed that Flexible-beta clustering with beta value -0.1 had best performance and Flexible-beta clustering with beta value -0.25 and -0.4 were proper performance in Hyrcanian regions. Flexible-beta clustering with beta value -0.25 was superior to others and Flexible-beta clustering with beta value -0.1 had the second rank. Since, choosing the most suitable clustering method is critical for achieving maximally ecological interpretable results, therefore, we suggested using flexible beta clustering with beta value equal to -0.1 and -0.25 in the studies area.

Nowadays, the application of clustering methods is widely increased, although choosing the right method due to the existence of different method and effective factors is difficult. The present study aimed to compare the results of widely used clustering algorithms and to determine the most effective methods according to the different evaluators and evaluate the effective distance measurement method for clustering algorithms. The data of Hyrcanian beech forests were examined in an area protected by the department of natural resources of Nowshahr. Random-systematic sampling method with regular grid of 100×200 m was used for determining the center of sample plots; 100-m2 (10×10 m) sample plots had been used to check the shrub species and 400-m2 (20×20 m) to check the herbaceous species. A total of 120 sample plots were measured. The abundance and coverage of tree, shrub and herbaceous species were estimated based on Braun-Blanquette scale. Three distance methods of measuring distance Bray Curtis, Hellinger and Manhattan were used and five clustering methods (Average method clustering methods, Ward method, flexible beta method with beta values ​​of -0.1, -0.25, -0.4) with six evaluation indicators (silhouette evaluation criterion, PARATNA criterion, Indval criterion, ISAMIC criterion, MRPP criterion and Phi correlation coefficient) were examined. Different clustering algorithms were arranged from best to worst for each dataset. The comparison analysis revealed that Ward’s and flexible-beta with beta value of -0.1 had the best performance. The present findings illustrated that Hellinger distance measurement method is better in homogeneous data than other distance measurement methods.

In this study, to investigate the water quality of Gamasyab and Ghareh-Sou rivers in Kermanshah province, data from a 5-year statistical period during 2014-2018 was used. To evaluate the water hydrochemical properties, the water type and hydrogeochemical facies of rivers water were first determined using Stiff and Piper diagrams. Next, the controlling factors of the water chemistry of two rivers were determined using Gibbs diagram and Ionic ratios. Also, factor analysis and cluster analysis were used to determine the processes affecting the hydrochemistry of river water. Finally, to predict the possibility of dissolution and precipitation of some minerals, their saturation indices were estimated. The results showed that water type and facies are bicarbonate calcite in both rivers. Also, the main factor in changing the chemical quality of water in both Gamasyab and Ghareh-Sou rivers is water-rock reaction. According to the results of factor and cluster analysis, this factor can be attributed to the dissolution of (1) carbonate and evaporation (2) dolomite formations in the rivers watershed. However, the dissolution of carbonate formations due to their abundances and distributions two watersheds has a more significant effect on the change in water chemistry.This resulted an increase in some chemical parameters such as calcium, magnesium and bicarbonate in the water of these two rivers and also caused positive water saturation index for aragonite, calcite and dolomite minerals. Considering the low contribution of the second factor in changing the chemical parameters, the saturation indices for evaporative minerals of gypsum, halite and anhydrite is negative.

The lack of complete coverage of hydrological data forces hydrologists to use the homogenization methods in regional analysis. In this research, in order to choose the best Hierarchical clustering method for regional analysis, base flow and related index were extracted from daily stream flow data using two parameter recursive digital filters in 43 hydrometric stations of the Kerman province. Physiographic, climatic and hydrological parameters were calculated. Using factor analysis of 14 parameters, six factors including: hydrograph recession constant, mean annual precipitation, actual evapotranspiration, permeability, topography (slope and height) and curve number were chosen as the most important independent factors. Hierarchical clustering of catchments using independent factors in different methods including: Nearest neighbor, Furthest neighbor, Centroid clustering and Ward were down. Then the regional equations using linear regression at 1% significant level for each homogeny region were determined. To compare and evaluate the accuracy and efficiency of the models, Independence errors, Colinerity and normal distribution of errors were tested. The results showed that “Centroid clustering method " because of low relative error with the amount of 27.137% for zone A and 36.89% for zone B of homogeneous region, minimum standard error of the estimate and the high correlation coefficient, diagnosed as the best homogeneity method to estimate regional base flow index in the study area.

Soil salinity is considered as one of the potential environmental hazards. The purpose of this study was to find the best index and the most suitable relationship for estimating soil salinity and its mapping using remote sensing data. At the first step, random sampling was performed using fishnet method and surface soil electrical conductivity (EC) measurements. Then, the threshold levels (92%, 95%, and 98%) were applied to the output images of each indicator. The methodology included using the least squares fitting (LS-fit) technique and principal components analysis (PCA) for halite and gypsum minerals, determining the correlation between the output of indices and ground data, and performing clustering and factor analysis between EC and output images. In order to select the best model derived from Landsat-8 band combinations and the amount of salinity, collinearity test, Durbin-Watson test, and backward multivariate regression were employed. The Cohen‘s kappa coefficient was also applied to evaluate the multivariate regression formed by Landsat-8 bands. The performance of the indicators was evaluated based on four criteria of root mean square error (RMSE), mean bias error (MBE), mean absolute error (MAE) and R-squared (R2). The results of the factor analysis showed the smallest distance between the EC, salinity index (SI) and brightness index (BI). The SI with an amount of 0.89 had the highest Pearson correlation with EC. In the dendrogram diagram, SI index with EC was placed in a cluster, and the RMSE, MBE, MAE and R2 values of the SI index were estimated to be 0.16, 0.11, 0.12, and 0.76, respectively. Compared to the rest of the indicators and linear, multivariate regression (with Cohen‘s kappa coefficient of 60%,), the SI index has provided better outcomes.