Income Tax Fraud Detection Using Machine Learning

The predict parameters of tax fraud detection specialists and researchers have utilized various approaches in identifying suspicious cases; the current research paper applied a supervised machine-learning model such as Random Forest, Decision Tree, Artificial Neural Network, and Logistic Regression. Based on various model evaluation indicators, Artificial Neural Network proved to be the most stable model in fraud tax prediction, This research paper was an alternative to the previous research paper, which used Gradient Boosting algorithms under machine learning to identify potential instances of income tax fraud. It entails the compilation and uses of the income tax datasets for the training and testing purpose. The aim primarily focuses on how optimum fraud detection accuracy can be achieved by using the boosting algorithms and visualizing critical attributes that define fraud. The evaluation will rest upon key metrics of measurement that include accuracy, sensitivity, specificity, the aim is to automate fraud detection, accurately, it concentrates on the correct identification of the difference between legitimate and hence scam and false tax filings, hence minimizing errors, for efficient and secure incomes tax assessment