![Dr. Igor Pantić](http://opensz.oss-cn-beijing.aliyuncs.com/Photo/2437195cfbec49738cd7595654996ec6.jpg)
Dr. Igor Pantić
ProfessorDepartment of medical physiology, Faculty of Medicine, University of Belgrade, Serbia
University of Haifa, Israel
College of Medicine and Health Sciences, Khalifa University of Science and Technology, UAE
Speech Title: Artificial intelligence approaches based on Support vector machine models in contemporary medical physiology research
Abstract: Support Vector Machines (SVM) are a class of supervised machine learning algorithms that can be used for predictive modeling in various fundamental and applied medical sciences. It is commonly used to create a hyperplane that differentiates the data into categories and this differentiation can be achieved either by linearly separating the data or by projecting it into a higher-dimensional space. Such techniques have promising applications in the classification of data obtained from signal analysis in medical physiology and related fields. In the past, various SVM models have been developed to evaluate patterns in heart rate variability, ECG signals, and to predict the outcomes of certain cardiovascular diseases. Large amounts of data from gene expression profiles can be harnessed for SVM training and testing, subsequently producing a model that contributes to predicting disease susceptibility. Support vector machines can also be employed to automate and expedite contemporary research protocols in experimental physiology. In this paper, we present several sophisticated SVM models developed in our previous research for the evaluation of two-dimensional signals. We also focus on the limitations of SVM models in medical physiology and issues related to their interpretability and reproducibility.
Keywords: Machine learning, Signal, Cell, Physiology, Algorithms, Predictive Modeling
Acknowledgements
This research was supported by the Science Fund of the Republic of Serbia, grant No. 7739645 “Automated sensing system based on fractal, textural and wavelet computational methods for detection of low-level cellular damage”, SensoFracTW.
Biography: Prof. Igor Pantić is an experienced researcher ranked among the top 2% of the most influential world scientists by Stanford University in 2021. He holds the rank of Associate Professor at the University of Belgrade, Faculty of Medicine (UBFM), and is currently employed at the Department of Medical Physiology. From 2017, he is also an Affiliated professor of the University of Haifa, Israel and from 2022 a Visiting Associate Professor at the Ben-Gurion University of the Negev, Faculty of Health Sciences, Department of Physiology and Cell Biology. Prof. Pantić is a data scientist with interest in the fields of machine learning and artificial intelligence and their applications in medicine. He is the Head of the Laboratory for cellular physiology at the UBFM and the author of numerous research articles in prestigious international scientific journals in the fields of molecular medicine, computational biology and neurosciences.