We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Cuckoo Search based Ensemble Classifier for Predictive Analysis of Malaria Infection Scope on thin Blood Smears.
- Authors
Jagtap D., Chaya; Rani N., Usha
- Abstract
Classification of machine learning models had an ultimate achievement by means of supervised learning, but the "state of the art models" have not yet extensively applied the "biological image data". To categorize the erythrocytes as malaria infested or not, we categorize erythrocytes thru ensemble classification method. With this regard, the script strived to prolong our former "decision tree based binary classifier" to accomplish the ensemble classification. Training data which is given clustered into diverse groups that are based on the variety perceived in all probable features projection. Every cluster is engaged to a greater extent to train single classifier. In this esteem, morphological features and entropy-based features are included. From legitimate pathology laboratories, the investigation carried on the real time inputs and combination of benchmark datasets composed as anonymous data. Unlike the existing methods, proposal of this document executed experiments on voluminous data that proliferate in size when compared to present benchmark datasets. With the statistical assessment the execution of proposal is evaluated by comparative analysis amid the two existing models "Malaria infected erythrocyte classification based on a hybrid classifier using microscopic images of thin blood smear (Hybrid Classification Approach)" and "Scale to Estimate Premature Malaria Parasites Scope (SEMPS) "and the proposed model "Cuckoo Search Based Ensemble Classifier (CSEC)". As depicted by the experimental study the suggested model is surpassing the two existing models.
- Publication
Indian Journal of Public Health Research & Development, 2019, Vol 10, Issue 5, p1019
- ISSN
0976-0245
- Publication type
Academic Journal
- DOI
10.5958/0976-5506.2019.01209.9