Handling Autism Imbalanced Data using Synthetic Minority Over-Sampling Technique (SMOTE)

被引：0

作者：

El-Sayed, Asmaa Ahmed ^{[1
]}

Meguid, Nagwa Abdel ^{[2
]}

Mahmood, Mahmood Abdel Manem ^{[1
]}

Hefny, Hesham Ahmed ^{[3
,4
]}

机构：

[1] Cairo Univ, Comp Sci, Dept Comp Sci, ISSR, Cairo, Egypt

[2] Natl Res Ctr, Human Genet, Res Children Special Needs Dept, Cairo, Egypt

[3] Cairo Univ, Comp Sci, ISSR, Cairo, Egypt

[4] Cairo Univ, ISSR, Cairo, Egypt

来源：

PROCEEDINGS OF 2015 THIRD IEEE WORLD CONFERENCE ON COMPLEX SYSTEMS (WCCS) | 2015年

关键词：

autism spectrum disorder (ASD); autism diagnostic interview-revised (ADI-R); Over-sampling; machine learning; support vector machine; Decision tree; navebayes; Cross-Validation;

D O I：

暂无

中图分类号：

TP301 [理论、方法];

学科分类号：

081202 ;

摘要：

The autism diagnostic interview-revised (ADI-R) is a semi-structured interview designed to assess the three core aspects of autism spectrum disorder (ASD). In this research a synthetic minority over-sampling technique (SMOT) was presented for handling autism imbalanced data to increase accuracy credibility. SMOT can potentially lead to over fitting on multiple copies of minority class examples. The autism data collected from National Research Center in Egypt (NRC). The experimental dataset applied on several machine learning algorithms and compared the accuracy before and after over-sampling techniques. The result show that over-sampling for imbalanced data making accuracy realistic and non-deceptive and can be Reliable.

引用

页数：5

共 13 条

[1] [Anonymous], 1997, Machine Learning
[2] [Anonymous], 2012, PUNE INT J EMERGING, V2
[3] Chawla N.V., SMOTE SYNTHETIC MINO
[4] Chawla Nitesh V., DATA MINING IMBALANC
[5] Chawla V., 2002, J ARTIFICIAL INTELLI, V16
[6] Fernandez A, 2008, FUZZY SETS SYSTEMS, V159
[7] Fernndez Alberto, 2009, EXPERT SYSTEMS APPL, V36
[8] Hall A., FEATURE SELECTION DI
[9] Hlosta Martin, 2013, INT J MACHINE LEARNI, V3
[10] Ishibuchi H., 2005, IEEE T FUZZY SYST, V13

← 1 2 →