No Detail Found!

Optimizing decision tree criteria for predicting covid-19 mortality in south korea dataset

Author(s)
IVAN DIRYANA SUDIRMAN Entrepreneurship Department, BINUS Business School Undergraduate Program, Bina Nusantara University, Bandung Campus, Bandung, Indonesia, 40181
RUDY ARYANTO Entrepreneurship Department, BINUS Business School Undergraduate Program, Bina Nusantara University, Bandung Campus, Bandung, Indonesia, 40181
MULYANI Entrepreneurship Department, BINUS Business School Undergraduate Program, Bina Nusantara University, Bandung Campus, Bandung, Indonesia, 40181
Abstract
SARS CoV 2 spreads very quickly. When this research was written, there were 1,780,315 positive cases globally. Countries in the world take various actions to curb the spread of this new virus. China is taking strict lockdown steps, while other countries such as South Korea are using massive diagnoses. While it might only be acceptable for countries with not too big populations, what the South Korean Government is doing is very interesting to research so that it can be recognized by other countries. This study use data mining techniques to search for pattern or new information from the data of the Covid-19 patients in South Korea. This study studies the use of decision trees to process the data in South Korea by finding the most optimal decision tree criteria. This study contributes to the data mining related field by showing decision tree is good enough to analyze the data. There are three best criteria for using a decision tree to predict deaths from this new virus. The best decision tree metrics are the gain ratio, knowledge gain and the gini index. In this analysis the gain-ratio is use for further analysis. Another finding is the province or location of this virus is the important factor. The implication of this finding supports the strategy of reducing virus transmission. Sex and age also play an important role in the prediction model.
Publication Details
Page(s) 2889-2900
DOI DOI not available
Published Journal: Journal of Theoretical and Applied Information Technology, Volume: 98, Issue: 15, Year: 2020

Request the full text from PASTIC

Keywords
Classification, Data Mining, decision tree, COVID-19, CRISP-DM
0

Citations

0

Downloads

25

Views