Decision Templates based Ensemble Classifiers for Automobile Insurance Fraud Detection
Slokashree Padhi, Suvasini Panigrahi · 2019 Global Conference for Advancement in Technology (GCAT) · 2019
In this paper, we have removed the class imbalance problem using SMOTE and used Decision Templates ensemble technique for efficiently detecting the auto insurance fraud. We employed three supervised classifiers namely, Support Vector Machine, Multilayer Perceptron and K-nearest Neighbors for classification purpose. The final classification results are obtained by aggregating the results obtained from the classifiers using Decision Templates technique. Our model has been experimentally evaluated with a real -world automobile insurance dataset.