Identification of Potential Quality of Groundwater Using Improved Fuzzy C Means Clustering Method

Abstract

The groundwater quality assessment gained more attention among the water quality management stations and researchers. The conventional water quality index method and artificial neural network models are used to assess groundwater. But these models are inadequate to handle data with uncertainty. In this work, we propose an improved Fuzzy C Means clustering method to identify the homogeneous clusters with respect to groundwater quality. For this purpose 1020 groundwater samples data with 7 physiochemical parameters of the year 2019 are collected from West Godavari, Andhra Pradesh, India. The effectiveness of the proposed clustering method is evaluated with two standard clustering methods namely K-means and Fuzzy C Means. The initial selection of the number of clusters and cluster centers determines the success of both the conventional K Means and Fuzzy C Means clustering methods. The proposed improved Fuzzy C Means method identifies the optimal number of clusters based on the water index value. The proposed improved Fuzzy C Means clustering method is implemented on the groundwater data set. The performance is computed with the help of the silhouette score and Davies Bouldin Index. The proposed clustering method outperforms with the existing K Means and Fuzzy C Means with silhouette score of 0.857 and Davies Bouldin Index value of 0.502 when the number of clusters are 4.

Block-Based K-Medoids Partitioning Method with Standardized Data to Improve Clustering Accuracy

“…Pre-processing is one of the important stages in data analysis which often improves the quality of the results of a method [12][13][14][15][16]. In cluster analysis, pre-processing can be done with transformations to standardize data [12,13,16]. This paper uses Eq.…”