期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2021
卷号：99
期号：22
语种：English
出版社：Journal of Theoretical and Applied
摘要：Recently, various uses of supervised classification recognition algorithms for medical images are reported in literature. Specifically, in the current deep learning era, machine learning techniques are considered as the most important and used approach for automatic healthcare systems. In this context, many comparisons of supervised deep learning techniques, more precisely, the neural one, are proposed. The proposed approach provides a medical assistance based on relevant aspects of Machine -Learning methods applied for blood cells objects recognition while taking into consideration the property of uncertainty of this kind of image. The overview presented in this article examines the existing literature and the contributions already done in the field of intelligent healthcare systems for blood cell images classification. For this purpose, we summarize previous efforts made to define recognition process with supervised deep learning method, establishing a novel definition of personalized Machine- Learning with a major focus on the uncertainty input image. Departing from this definition, we propose and discuss the efficiency of Convolutional Neural Network for which the architecture is built and examined in detail. A Bayesian optimization of Convolutional Neural Network hyper parameters is also proposed. The main goal is to increase recognition rate while respecting time complexity. That is why an experimental comparison of Convolutional Neural Network with Support Vector Machine and K- nearest neighbor performance is discussed.
