Identification0of Leukemia0Subtypes0Using Convolutional0Neural0Network

  • P.Rajesh Kanna, S.Kirubakaran, R.Mani Bharathi, S.Rajaram, J.Sam Joesph

Abstract

Leukemia0 is a deadly malignant growth and has two principle types: Acute and ceaseless. Each type has two more subtypes: Lymphoid and myeloid. Consequently, altogether, there are four subtypes of leukemia. We proposes another methodology for determination of all subtypes of leukemia from minuscule platelet pictures utilizing convolutional neural systems 0(CNN), which requires an enormous preparing informational collection. Along these lines, we likewise researched the impacts of information growth for an expanding number of preparing tests artificially. A convolution neural systems (CNN) design equipped for perceiving all subtypes of leukemia. Additionally, we likewise investigated other understood AI calculations, for example, guileless Bayes,0bolster vector machine, k-closest neighbor, and choice tree. The outcomes we acquired from tests demonstrated that our CNN0 model execution has more exactness, in leukemia versus solid and multi-class order of all subtypes, individually. At long last, we likewise demonstrated that the CNN0 model has a superior exhibition than other understood AI calculations.

Published
2020-03-30
How to Cite
P.Rajesh Kanna, S.Kirubakaran, R.Mani Bharathi, S.Rajaram, J.Sam Joesph. (2020). Identification0of Leukemia0Subtypes0Using Convolutional0Neural0Network. International Journal of Advanced Science and Technology, 29(4s), 2120 - 2127. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/8414