Healthcare insurance fraud detection using data mining

Table 1 Study Methods and Features for Work on DE-synPUF dataset

Study reference	Labeling approach	Techniques used
Bauder et al. [44, 45]	LEIE	Random Forest, C4.5, SVM, Logistic Regression
Herland et al. [46, 47]	LEIE	Logistic Regression, Gradient Tree Boosting
Fan et al. [48]	LEIE	Logistic Regression, Naïve Bayes, Decision Tree
Ekin et al. [49]	Unsupervised	PCA, RWO, co-clustering
Sadiq et al. [50] (2017)	PRIM	bump hunting
Sadiq et al. [51] (2019)	unsupervised	Cascaded Propensity Matching (CPM) Fraud Miner
Zafari and Ekin [52]	Unsupervised	Topic modeling, outlier detection
Ekin et al. [53] (2019)	Unsupervised	Bayesian model, Gibbs sampling

ISSN: 1472-6947