Thanjida Akhter
M.Sc. Candidate
Supervisor: Dr. Ting Hu

Implementation of machine learning algorithm to find classification of features of osteoarthritis (OA)

Department of Computer Science
Wednesday, September 12, 2018, 12:00p.m., EN-2022

Abstract

Machine learning explores the study and construction of algorithm that can learn from and make predictions on data, classification of samples or features, development of metabolic networks, or properties of system, or to develop, augment or verify models and as well as the analysis of major features. Osteoarthritis (OA) is a heterogeneous disease with various pathogenic factors and a most common form of arthritis. OA have huge metabolic statues, if we are able to find those Important metabolic then we can predict the disease. Three Machine learning algorithm were used in this project to find those Important metabolic of OA rank by each algorithm.

In this project, three Machine learning algorithm were used are fandom Forest, Gradient Boosting Machine (GBM) and K-nearest Neighbors (k-NN) can show the classification of features of osteoarthritis (OA). First, we represent the accuracy of three algorithm how they predict a disease and non-disease data. Using cross-validation for more accurate result and then show top eleven (11) metabolic by each algorithm. Three algorithm have some common metabolic. Finally, in this project twenty (20) Important metabolic of OA were found from all three algorithm.