TY - GEN
T1 - Performance improvement of decision trees for diagnosis of coronary artery disease using multi filtering approach
AU - Abdar, Moloud
AU - Nasarian, Elham
AU - Zhou, Xujuan
AU - Bargshady, Ghazal
AU - Wijayaningrum, Vivi Nur
AU - Hussain, Sadiq
N1 - Publisher Copyright:
© 2019 IEEE.
PY - 2019/2
Y1 - 2019/2
N2 - The heart is one of the strongest muscular organs in the human body. Every year, this disease can kill many people in the world. Coronary artery disease (CAD) is named as the most common type of heart disease. Four well-known decision trees (DTs) are applied on the Z-Alizadeh Sani CAD dataset, which consists of J48, BF tree, REP tree, and NB tree. A multi filtering approach, named MFA, was used to modify the weight of attributes to improve the performance of DTs in this study. The model was applied on three main coronary arteries including the Left Anterior Descending (LAD), Left Circumflex (LCX), and Right Coronary Artery (RCA). The obtained results show that data balancing has a valuable impact on the performance of DTs. The comparison results show that this study provides the best results applied on the Z-Alizadeh Sani dataset compared to previous studies. The proposed MFA could improve the performance of the classic DTs algorithms significantly, with the highest accuracies obtained by NB tree for LAD, LCX, and RCA are 94.90%, 92.97% and 93.43%, respectively.
AB - The heart is one of the strongest muscular organs in the human body. Every year, this disease can kill many people in the world. Coronary artery disease (CAD) is named as the most common type of heart disease. Four well-known decision trees (DTs) are applied on the Z-Alizadeh Sani CAD dataset, which consists of J48, BF tree, REP tree, and NB tree. A multi filtering approach, named MFA, was used to modify the weight of attributes to improve the performance of DTs in this study. The model was applied on three main coronary arteries including the Left Anterior Descending (LAD), Left Circumflex (LCX), and Right Coronary Artery (RCA). The obtained results show that data balancing has a valuable impact on the performance of DTs. The comparison results show that this study provides the best results applied on the Z-Alizadeh Sani dataset compared to previous studies. The proposed MFA could improve the performance of the classic DTs algorithms significantly, with the highest accuracies obtained by NB tree for LAD, LCX, and RCA are 94.90%, 92.97% and 93.43%, respectively.
KW - Classification
KW - Coronary artery disease
KW - Data mining
KW - Heart disease
KW - Machine learning
UR - http://www.scopus.com/inward/record.url?scp=85072981731&partnerID=8YFLogxK
UR - https://iccc2019.ieee-iccc.org/
U2 - 10.1109/CCOMS.2019.8821633
DO - 10.1109/CCOMS.2019.8821633
M3 - Conference contribution
AN - SCOPUS:85072981731
SN - 9781728113234
T3 - 2019 IEEE 4th International Conference on Computer and Communication Systems, ICCCS 2019
SP - 26
EP - 30
BT - 2019 IEEE 4th International Conference on Computer and Communication Systems, ICCCS 2019
A2 - Xiao, Yang
A2 - Funabiki, Nobuo
PB - IEEE, Institute of Electrical and Electronics Engineers
CY - United States
T2 - 4th IEEE International Conference on Computer and Communication Systems, ICCCS 2019
Y2 - 23 February 2019 through 25 February 2019
ER -
