Classification of white blood cell images using K-Medoids algorithm and comparison of accuracy in terms of CNN technique

In the proposed research, white blood cell images will be classified using the K-Medoids method and the CNN algorithm will be compared. There are 790 photos in the dataset-master image dataset, on which K-Medoids is used. For the categorization of white blood cell pictures, a Deep learning approach...

Full description

Saved in:
Bibliographic Details
Main Authors: Latha, Nuka Pushpa, Senthilkumar, R.
Format: Conference Proceeding
Language:English
Subjects:
Algorithms
Artificial neural networks
Classification
Datasets
Error correction
Leukocytes
Machine learning
Pictures
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In the proposed research, white blood cell images will be classified using the K-Medoids method and the CNN algorithm will be compared. There are 790 photos in the dataset-master image dataset, on which K-Medoids is used. For the categorization of white blood cell pictures, a Deep learning approach that compares Convolutional Neural Network with K- Medoids has been suggested and developed. It was determined that each group had a sample size of 27 people. The categorization of pictures of blood cells was examined and documented for its correctness and sensitivity. When compared to a Convolutional Neural Network, K-Medoids classified blood cell pictures with the highest accuracy (91.8 percent) and the lowest mean error (86.4 percent). The classifiers have a significant difference of 0.05. K-Medoids Algorithm outperforms Convolutional Neural Network in the classification of blood cell pictures, according to a new research.
ISSN:0094-243X
1551-7616
DOI:10.1063/5.0228283