A study of agglo hi clustering techniques for gene expression micro array data in data mining

Abstract

Every single living organism shares numerous attributes in common from generation to generation. The recognizable attributes of organisms like color, size, and shape are inherited by the next generation from the parent. The genes exist in all living beings. The genes reside throughout the body of all living beings. Genes are a set of instructions that determine what the living being is like, its look, how it stays alive and the change of behavior with the environment. The Gene expression is the process by which the instructions in the Genes are converted into a functional product, such as a protein through the functional gene product synthesis. The Gene expression basically encodes the proteins which in turn command the cell functionality. Gene expression is the process in which the heritable information present inside a gene is made into a functional gene product called protein or Ribonucleic acid. The Gene expression data represents a condition matrix where each row corresponds to the gene and the column signifies the condition. Each element in the matrix is a real number and records the expression level of a gene under a specific condition. The organic system is very complex and the genes volume also increases in the biological networks that leads to the difficulties in the Gene data interpretation which itself inhibit vagueness, noise and imprecision. Clustering is an unsupervised learning technique which applies statistical data analysis to identify patterns and make decision. The clusteringis made in effective manner by applying suitable technique with proper parameters like distance function, threshold values etc. Data pre-processing has to carried out and the model parameters are modified until the desired results are achieved. Clustering methods will help to expose the structures and the patterns in the original data for taking further decisions. The traditional clustering techniques provide a reliable output in clustering, but they fail to deal with scalability and they also fail to identify the number