Publications

Abstract : In the medical field to solve the new issues, the novel approaches for managing relevant features by using genomes are considered; using the sub-sequence of genes the outcome of interest is analyzed. In this implementation part of the model, we have given the MEDLINE and PubMed archives as inputs to the proposed model. A large number of MESH terms with gene and protein are utilized to characterize the patterns of a large number of medical documents from a large set of records. Standard datasets with different characteristics are used for examination study. The characteristics and inadequacies of different techniques are noted. Feature selection techniques are given in perspective of data composes and region traits by applying proper rules. Feature context extraction through name element distinguishing proof is an essential errand of online therapeutic report grouping for learning disclosure databases. The parameters are identified to compare with other models implemented on these datasets and the results prove that the proposed method is very effective than existing models. The primary point of the proposed ensemble learning models is to characterize the high dimensional information for gene/protein-based disease expectation in light of substantial biomedical databases. The proposed model uses an efficient ranking algorithm to select the relevant attributes from a set of all attributes; the attributes are given to the classifier to improve the accuracy based on the users’ interest.