DOI QR코드

DOI QR Code

Operation Plan of Big Data Prediction Model using Cut-off-Voting Classifier in Administrative Big Data Environment

행정 빅데이터 환경에서 컷오프-투표 분류기를 활용한 빅데이터 예측모형의 실험

  • Woosik Lee (Research Center, Korea Social Security Information Service)
  • 이우식 (한국사회보장정보원 사회보장정보연구소)
  • Received : 2024.03.04
  • Accepted : 2024.04.20
  • Published : 2024.05.31

Abstract

In order to operate predictive models utilizing administrative big data, it is crucial to consider policy changes and the characteristics of highly volatile data. Considering this scenario, this study proposes the Cut-off Voting Classifier (CVC) algorithm. This proposed algorithm prevents a sharp decline in accuracy by utilizing multiple weak classifiers. The study validates the proposed algorithm's performance through experiments. The performance evaluation demonstrates the ability to maintain stable prediction rates even in situations with a sharp decline in predictive model accuracy.

행정 빅데이터를 활용하는 예측 모형을 운영하기 위해서는 정책의 변화 및 변동성 심한 데이터의 특성이 고려가 되어야만 한다. 이런 상황을 고려하여 본 연구에서는 Cut-off Voting Classifier(CVC) 알고리즘을 제안한다. 제안하는 알고리즘은 여러개의 약 분류기를 활용하여 적중률이 급격하게 하락하는 것을 방지하는 알고리즘이다. 본 연구에서는 제안하는 알고리즘을 실험을 통해 성능을 검증한다. 성능검증 결과 급격하게 예측모형 적중률이 하락하는 상황에서도 안정적으로 예측률을 유지한다는 것을 입증할 수 있었다.

Keywords

References

  1. Yoo Jong-seong, Jeon Byung-yu, Shin Kwang-young, Lee Do-hoon, Choi Seong-su. Utilization of administrative data for evidence-based policy research. Korea Social Policy Review. 2020;27(1);5-37
  2. Elias, P. Administrative data: Facing the future: european research infrastructures for the humanities and social sciences, SCIVERO. 2014:47-48.
  3. Woollard, M. Administrative data: problems and benefits. A perspective from the United Kingdom. Facing the future: european research infrastructures for the humanities and social sciences, SCIVERO. 2014.
  4. Big Data Analysis and Utilization Division, Ministry of Public Administration and Security. New design of public service based on public experience with data - Confirmation and announcement of 「First Basic Plan for Data-Based Administration Revitalization (2021~2023)」 1st meeting held-. Ministry of Public Administration and Security, 2021.2.20.
  5. Cha Kyung-yeop, "A study on the development of a model for predicting illegal receipt of the National Pension using data mining - for infidelity reporting of damages -. Proceedings of the Korean Statistical Society," 2010:17(1):1-8 https://doi.org/10.5351/CKSS.2010.17.1.001
  6. Kim Young-sun, Park Seon-mi, Choi Ki-jung, Park Eun-ang, "Measures to improve social service abnormal payment and illegal supply and demand monitoring system," Korea Social Security Information Service. 2017.12.
  7. Na Young-gyun, Cha Ye-rin, Choi Dae-gyu, ImSeung-ji, Kim Na-young, "A study on improvement of health insurance arrears collection. Health Insurance Corporation," 2020.12.
  8. 6494 cases of illegal subsidies caught by AI of the Ministry of Strategy and Finance... Electronic newspaper with list confirmed by May, 2021.03
  9. Prevention of Unauthorized Receiving of Unemployment Benefit-Public Big Data Part 5-. Ministry of Public Administration and Security, 2017.03.
  10. Expansion of crime prevention activities using police big data and artificial intelligence (AI) nationwide. Police Agency Press Release, 2021.04
  11. Yujin Gil, Yoon Chung and Sangsoo Park, "Diabetes Prevalence and Diagnosis Rates, and Risk Factor Effect Analysis," JCCT, 2024.1.31.
  12. Seoul Life Movement Data Manual. Seoul City. Korea Transportation Research Institute, KT, 2021.09
  13. Hye-Sun Lee, "Study of mental disorder Schizophrenia, based on Big Data," IJACT, 2023.11
  14. F. Mutaher and Ba-Alwi, Comparative Study for Analysis the Prognostic in Hepatitis Data: Data Mining Approach, International Journal of Scientific and Engineering Research, 2013.8.