• Malik Tajamul Hussain

Over the years, several data mining techniques have been designed. There are few traditional statistical methods applied to provide more reliable and scalable tools for making improvements. Since data mining is applied in many applications, improvements are being made with time. Clustering is known as the method using which the data is classified into groups based on its similarity patterns. Simplification is achieved in case when minimum numbers of clusters are included. However, in case of less numbers of clusters, some file details are lost. Clusters help in performing data modeling. The clusters can be considered as hidden patterns as per the machine learning view and the unsupervised method can be used to search content. A data concept is defined by a system that provides an output. The clustering method is not based on one step which is stated in its definition. The type of clustering that defines the clusters based on the density of data is known as density based clustering. The density based algorithms help in discovering the arbitrary shaped clusters which results in providing an authentication against the outliers entering the systems. The density based algorithms are responsible for solving the issues related to partition relocation clustering that are faced by clusters. Another property identified here is good scalability. Along with the benefits, the technology also shows some drawbacks. For applying density-based algorithms, a metric space is required. A natural setting provided here is the spatial data clustering. The density based algorithms require a metric space and the spatial data clustering is performed through natural setting. One of the most commonly used bio-inspired algorithms of this research is Particle Swarm Optimization (PSO). The artificial neural networks are trained using this method. An error value is calculated for each neuron present in the output layer. Initiating from the output, backward propagation of error values is calculated. Improving the performance of incremental DBSCAN is the objective of this research. Based on the input dataset the EPS value is defined in incremental DBSCAN algorithm. By reducing the accuracy of clustering, it is possible to keep the value of Euclidean distance static. The PSO algorithm is used in this research to improve the performance of incremental DBSCAL algorithm. Here, the Euclidian distance is calculated in dynamic manner and due to which the execution time of clustering is reduced thus, increasing its accuracy. Each point and their value will be taken by the PSO as input and at every clustering point the error will be calculated. The specific point at which accuracy of clustering is the highest is considered as the best point since at this point, the error is the least. The accurate point for clustering is defined by the efficient calculation of Euclidian distance. The similarity among data points for clustering is defined by distance.

"Density Based Clustering Method Using Bio-Inspired Method", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 6, page no.840-844, June 2019, Available :http://www.jetir.org/papers/JETIR1908125.pdf

"Density Based Clustering Method Using Bio-Inspired Method", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 6, page no. pp840-844, June 2019, Available at : http://www.jetir.org/papers/JETIR1908125.pdf