Duma, MTwala, BNelwamondo, Fulufhelo VMarwala, T2013-10-232013-10-232013-09Duma, M, Twala, B, Nelwamondo, F.V and Marwala, T. 2013. Partial imputation to improve predictive modelling in insurance risk classification using a hybrid positive selection algorithm and correlation-based feature selection. Current Science, vol. 103(6), pp 697-7050011-3891http://www.currentscience.ac.in/Volumes/103/06/0697.pdfhttp://hdl.handle.net/10204/6984Copyright: 2013 Indian Academy of Sciences. This an open access journal. This journal authorizes the publication of the information herewith contained. Published in Current Science, vol. 103(6), pp 697-705We propose a hybrid missing data imputation technique using positive selection and correlation-based feature selection for insurance data. The hybrid is used to help supervised learning methods improve their classification accuracy and resilience in the presence of increasing missing data. The positive selection algorithm searches for potential candidates for imputation and the correlation-based feature selection method searches for attributes have a significant effect on the target outcome. The imputation is performed only on those attributes that have an impact on the target outcome. The results show that the classification accuracy and resilience of supervised learning methods improve significantly when applied with the imputation strategy under these assumptions.enInsurance risk classificationMissing dataPositive selection algorithmSupervised learning methodsInsurance dataArticleDuma, M., Twala, B., Nelwamondo, F. V., & Marwala, T. (2013). Partial imputation to improve predictive modelling in insurance risk classification using a hybrid positive selection algorithm and correlation-based feature selection. http://hdl.handle.net/10204/6984Duma, M, B Twala, Fulufhelo V Nelwamondo, and T Marwala "Partial imputation to improve predictive modelling in insurance risk classification using a hybrid positive selection algorithm and correlation-based feature selection." (2013) http://hdl.handle.net/10204/6984Duma M, Twala B, Nelwamondo FV, Marwala T. Partial imputation to improve predictive modelling in insurance risk classification using a hybrid positive selection algorithm and correlation-based feature selection. 2013; http://hdl.handle.net/10204/6984.TY - Article AU - Duma, M AU - Twala, B AU - Nelwamondo, Fulufhelo V AU - Marwala, T AB - We propose a hybrid missing data imputation technique using positive selection and correlation-based feature selection for insurance data. The hybrid is used to help supervised learning methods improve their classification accuracy and resilience in the presence of increasing missing data. The positive selection algorithm searches for potential candidates for imputation and the correlation-based feature selection method searches for attributes have a significant effect on the target outcome. The imputation is performed only on those attributes that have an impact on the target outcome. The results show that the classification accuracy and resilience of supervised learning methods improve significantly when applied with the imputation strategy under these assumptions. DA - 2013-09 DB - ResearchSpace DP - CSIR KW - Insurance risk classification KW - Missing data KW - Positive selection algorithm KW - Supervised learning methods KW - Insurance data LK - https://researchspace.csir.co.za PY - 2013 SM - 0011-3891 T1 - Partial imputation to improve predictive modelling in insurance risk classification using a hybrid positive selection algorithm and correlation-based feature selection TI - Partial imputation to improve predictive modelling in insurance risk classification using a hybrid positive selection algorithm and correlation-based feature selection UR - http://hdl.handle.net/10204/6984 ER -