Morehead State University
Classification of Stroke Victims Through Ensemble Learning and an Exploration of Various Supervised Machine Learning Algorithms
Grade Level at Time of Presentation
Sophomore
Major
Computer Science
Minor
Mathematics
Institution
Morehead State University
KY House District #
93
KY Senate District #
31
Faculty Advisor/ Mentor
Heba Elgazzar, PhD
Department
Computer Science
Abstract
Throughout the developed and developing world, a malady has plagued humanity since antiquity: stroke. This research project explores several methodologies within machine learning that enable the classification of two distinct classes: absence or presence of a history of stroke. This paper explores several contemporary methodologies for classifying stroke victims using an aggregate of ensemble learning and a myriad of other supervised learning strategies. To this end, the research presented in this paper makes explicit use of bagging and boosting techniques to strengthen classification efficacy. Several supervised learning algorithms are utilized to perform the classification process. Python programming language with the machine learning libraries were used to design classification models and generate illustrative graphics. The experimental results as discussed in greater detail in this paper, show that conditions such as stroke can be effectively classified with considerable high accuracy and precision.
Classification of Stroke Victims Through Ensemble Learning and an Exploration of Various Supervised Machine Learning Algorithms
Throughout the developed and developing world, a malady has plagued humanity since antiquity: stroke. This research project explores several methodologies within machine learning that enable the classification of two distinct classes: absence or presence of a history of stroke. This paper explores several contemporary methodologies for classifying stroke victims using an aggregate of ensemble learning and a myriad of other supervised learning strategies. To this end, the research presented in this paper makes explicit use of bagging and boosting techniques to strengthen classification efficacy. Several supervised learning algorithms are utilized to perform the classification process. Python programming language with the machine learning libraries were used to design classification models and generate illustrative graphics. The experimental results as discussed in greater detail in this paper, show that conditions such as stroke can be effectively classified with considerable high accuracy and precision.