Abstract:The advancement in healthcare services has been increasing widely to extend several services with intense quality. One of the important issues affecting the effective use of public funds is the detection of health insurance fraud. Previous techniques of detecting fraud pay close attention to characteristics of a single visit rather than many patient visits. Due to a higher false positive rate and poor profile construction, the common traits have reduced detection performance. This paper introduces a novel and intelligent Provider Fraud_Anomaly Detection System (PF_ADS) by combining big data and deep learning approaches for the healthcare insurance industry. The proposed framework contributes to improvising the preprocessing and classification phases to detect provider fraud at an untimely phase. Initially, the collected datasets are preprocessed using a Relative Risk-based MapReduce framework that builds an organized set of relationships between diseases, patients, and claiming variables. The classification phase is improvised using a proposed Recurrent Neural Network (RNN). It consists of sophisticated steps to consider the significant attributes using hyperparameter optimization. Recalling ability is one of the best parts of RNNs that defines the past and present states of the networks. Therefore, the ability of network state predictions and the tuning of parameters is studied by improved Decisional Score-based Bayesian Optimization (DS_BO). Finally, the best attributes with the selective hyperparameters are fed into the input layer of the Recurrent Neural Networks (RNNs) to classify the anomalies from the provider's end. The proposed PF_ADS framework is experimented with and validated on the public repositories. The experimental results state that the proposed framework outperforms better than the other methods in terms of accuracy (88.09%), precision (14.15%), recall (32.80%), and 92.30 s computational time.

Research on Bootstrapping Algorithm for Health Insurance Data Fraud Detection Based on Decision Tree.

Efficient fraud detection using deep boosting decision trees

Building prediction models and discovering important factors of health insurance fraud using machine learning methods

An Optimized LightGBM Model for Fraud Detection

An Ensemble Random Forest Algorithm for Insurance Big Data Analysis

Detection of Fraudulent Health Insurance Claims Based on Decision Tree with Principal Component Analysis

Fraud Detection Using Decision Tree Algorithm to Curb Identity Theft

Application of Clustering Methods to Health Insurance Fraud Detection

Exploring Maximum Tree Depth and Random Undersampling in Ensemble Trees to Optimize the Classification of Imbalanced Big Data

LightGBM Model for Detecting Fraud in Online Financial Transactions

Application of a Decision Tree Model for Predicting Diabetic Retinopathy

Health insurance fraud detection by using an attributed heterogeneous information network with a hierarchical attention mechanism

Detection of fraudulent users in P2P financial market

FraudAuditor: A Visual Analytics Approach for Collusive Fraud in Health Insurance

A Decision Tree Approach for Assessing and Mitigating Background and Identity Disclosure Risks.

Retrieval-Based Gradient Boosting Decision Trees for Disease Risk Assessment

Detection and Analysis of Credit Card Application Fraud Using Machine Learning Algorithms

Health Insurance Anomaly Detection Based on Dynamic Heterogeneous Information Network

Empirical Analysis of Financial Statement Fraud of Listed Companies Based on Logistic Regression and Random Forest Algorithm

Financial Fraud Detection: a New Ensemble Learning Approach for Imbalanced Data.

Design and development of big data-based model for detecting fraud in healthcare insurance industry