A Prediction Model for Blood Donation Using Multiple Logistic Regression

Authors

  • Wan Hanieza W. Mohamad Hanapi
  • Haslina Sarkan
  • Nilam Nur Amir Sjarif
  • Yazriwati Yahya
  • Suriayati Chuprat

Keywords:

Prediction model, machine learning, Logistic Regression Model

Abstract

Minimal participation in blood donation is a concern in this country despite many blood donation programs organized. This study attempts to find out the factors influencing the intention level of people to donate blood. The main objective of this paper is to identify the association between the willingness of donating blood with the number of months since the last donation, number of donation, total volume donated and the number of months since the first donation. Secondary data retrieved from UCI Machine Learning Repository were used. Based on the Logistic Regression Model, there are only three factors affecting the willingness to donate blood which are month since last donation, frequency of donation and total volume donated. There are four ways to validate the model which are Hosmer and Lemeshow Test, Coefficient of Determination, Classification Table and area under Receiver Operating Characteristics (ROC) value. The validation tests showed that the final model has a good performance.

 

 

Author Biographies

Wan Hanieza W. Mohamad Hanapi

 

 

Haslina Sarkan

 

 

Nilam Nur Amir Sjarif

 

 

Yazriwati Yahya

 

 

Suriayati Chuprat

 

 

Downloads

Published

2019-12-30

How to Cite

W. Mohamad Hanapi, W. H., Sarkan, H., Amir Sjarif, N. N. ., Yahya, Y., & Chuprat, S. . (2019). A Prediction Model for Blood Donation Using Multiple Logistic Regression. Open International Journal of Informatics, 7(Special Issue 2), 147–157. Retrieved from https://oiji.utm.my/index.php/oiji/article/view/87