Outlier Detection Using Association Rule Mining and Cluster Analysis

2018 ◽  
Vol 6 (6) ◽  
pp. 529-533
Author(s):  
C. Leela Krishna ◽  
C. Kala Krishna
Keyword(s):  
Cluster Analysis ◽  
Outlier Detection ◽  
Association Rule ◽  
Association Rule Mining ◽  
Rule Mining ◽  
And Cluster Analysis

Prevalence and Patterns of Multimorbidity in a Nationally Representative Sample of Older Chinese: Results From the China Health and Retirement Longitudinal Study

2019 ◽  
Vol 75 (10) ◽  
pp. 1974-1980 ◽  
Author(s):  
Shan-Shan Yao ◽  
Gui-Ying Cao ◽  
Ling Han ◽  
Zi-Shuo Chen ◽  
Zi-Ting Huang ◽  
...  
Keyword(s):  
Cluster Analysis ◽  
Longitudinal Study ◽  
Hierarchical Cluster Analysis ◽  
Association Rule ◽  
Association Rule Mining ◽  
Hierarchical Cluster ◽  
Urban Residents ◽  
Rule Mining ◽  
Nationally Representative ◽  
Older Chinese

Abstract Background Multimorbidity has become a prominent problem worldwide; however, few population-based studies have been conducted among older Chinese with multimorbidity. This study aimed to examine the prevalence of multimorbidity and explore its common patterns among a nationally representative sample of older Chinese. Methods This study used data from the China Health and Retirement Longitudinal Study and included 19,841 participants aged at least 50 years. The prevalence of individual chronic diseases and multimorbidity during 2011–2015 were evaluated among the entire cohort and according to residential regions and gender. The relationships between participants’ demographic characteristics and multimorbidity were examined using logistic regression model. Patterns of multimorbidity were explored using hierarchical cluster analysis and association rule mining. Results Multimorbidity occurred in 42.4% of the participants. The prevalence of multimorbidity was higher among women (odds ratio [OR] = 1.31, 95% confidence interval [CI]: 1.13–1.51) and urban residents (OR = 1.14, 95% CI: 1.02–1.27) than their respective counterparts after accounting for potential confounders of age, education, smoking, and alcohol consumption. Hierarchical cluster analysis revealed four common multimorbidity patterns: the vascular-metabolic cluster, the stomach-arthritis cluster, the cognitive-emotional cluster, and the hepatorenal cluster. Regional differences were found in the distributions of stroke and memory-related disease. Most combinations of conditions and urban–rural difference in multimorbidity patterns from hierarchical cluster analysis were also observed in association rule mining. Conclusion The prevalence and patterns of multimorbidity vary by gender and residential regions among older Chinese. Women and urban residents are more vulnerable to multimorbidity. Future studies are needed to understand the mechanisms underlying the identified multimorbidity patterns and their policy and interventional implications.

A Novel Market Basket Analysis Using Adaptive Association Rule Mining Algorithm

2012 ◽  
Vol 1 (4) ◽  
pp. 25-28
Author(s):  
M.Dhanabhakyam M.Dhanabhakyam ◽  
◽  
Dr.M.Punithavalli Dr.M.Punithavalli
Keyword(s):  
Association Rule ◽  
Association Rule Mining ◽  
Market Basket Analysis ◽  
Rule Mining ◽  
Market Basket ◽  
Mining Algorithm

Prediksi Code Defect Perangkat Lunak Dengan Metode Association Rule Mining dan Cumulative Support Thresholds

2015 ◽  
Vol 6 (2) ◽  
Author(s):  
Rizal Setya Perdana ◽  
Umi Laili Yuhana
Keyword(s):  
Association Rule ◽  
Association Rule Mining ◽  
Rule Mining ◽  
Program Code

Kualitas perangkat lunak merupakan salah satu penelitian pada bidangrekayasa perangkat lunak yang memiliki peranan yang cukup besar dalamterbangunnya sistem perangkat lunak yang berkualitas baik. Prediksi defectperangkat lunak yang disebabkan karena terdapat penyimpangan dari prosesspesifikasi atau sesuatu yang mungkin menyebabkan kegagalan dalam operasionaltelah lebih dari 30 tahun menjadi topik riset penelitian. Makalah ini akandifokuskan pada prediksi defect yang terjadi pada kode program (code defect).Metode penanganan permasalahan defect pada kode program akan memanfaatkanpola-pola kode perangkat lunak yang berpotensi menimbulkan defect pada data setNASA untuk memprediksi defect. Metode yang digunakan dalam pencarian polaadalah memanfaatkan Association Rule Mining dengan Cumulative SupportThresholds yang secara otomatis menghasilkan nilai support dan nilai confidencepaling optimal tanpa membutuhkan masukan dari pengguna. Hasil pengujian darihasil pemrediksian defect kode perangkat lunak secara otomatis memiliki nilaiakurasi 82,35%.

Sign in / Sign up

Export Citation Format

Share Document