Faramarz Karamizadeh ¹, Seyed Ahad Zolfagharifar ²

Affiliations
1 Department of Electrical and Computer Engineering, Shiraz University, Shiraz, Iran
2 Kohgiluyeh and Boyer Ahmad Science and Research Branch, Islamic Azad University, Iran

Abstract

Background/Objectives: Insurance data analysis can be considered as a way of losses reduction by using data mining. It uses the machine learning, pattern recognition and data base theory for discovering the unknown knowledge. Methods/Statistical Analysis: In this paper, information of 2011, third party insurance of Iran insurance company auto has analyzed in Kohgiluyeh and Boyer Ahmad by using the data mining method. Findings: The results show that using clustering algorithms with acceptable clusters will be able to provide a model to identify affecting factors and to determine the effect of them in the profit and loss of auto third party insurance. Applications/Improvements: The algorithm of K-Means has formed the best clustering with 9 clusters that have relatively good quality. It means that has been able to maximize the distance between the cluster and minimize the within cluster distance.

Keywords

Clustering Algorithm, Data Mining, Insurance, Profit and Loss, Third Party

Full Text

   

Seyed Ahad Zolfagharifar ¹, Faramarz Karamizadeh ²

Affiliations
1 Islamic Azad University Science and Research, Kohgiluyeh and Boyer Ahmad Branch, IR
2 Department of Electrical and Computer Engineering, Shiraz University, Shiraz, IR

Abstract

Objective: The objective of this paper is to give a hybrid classifier by combining the genetic algorithm and decision tree based on evolutionary learning. Methods: The proposed algorithm on the 8 data samples was tested. In order to implement the proposed algorithm, MATLAB software was used. In all the obtained results, standardized data sets are used, making assembly by using genetic algorithm which is very suitable. Results: The learning technique of sub-spaces is proposed. In this study, we tried to compare a series of different methods and updated of integrated distribution. It showed that, in cases that the number of information or the number of properties are high, the proposed hybrid classification approach that implements genetic algorithm can be used as the best approach. Conclusion: In this study, we tried a usual approach for clustering in error prone environments. A main excess in the precision on the tested information or on the validation is clear. It should be noted that this increasing is in comparison with the assembly classifiers which has stable accuracy.

Keywords

Combination of Genetic and Decision Tree, Consensus of Classifiers.

Full Text