Olamilekan002/Tunisian-Fraud-Detection-Challenge

Tunisian Fraud Detection Challenge by DSN AI+ Unilorin This repository is about an Hackathon Organized by Data Science Nigeria(DSN-AI) Unilorin. My 3rd place solution , Root Mean Square Error of 5.5229 on Zindi where the competition was hosted. Tax fraud is the intentional act of lying on a tax return form with the intent to lower one’s tax liability. Under-reporting is one of the most common types of tax frauds. It consists of filing a tax return form with a lesser tax base. As a result of this act, fiscal revenues are reduced, undermining public investment in much-needed services. The objective of the challenge is to detect tax fraud. This is one of the main priorities of local tax authorities which are required to develop cost-efficient strategies to tackle this problem. Using historical data, a supervised machine learning technique that detects potential fraudulent taxpayers will increase the operational efficiency of the tax supervision process. My Approach: Handled Missing Values, Drop a lot of Non important features, Robust Scaling, KFold Validation, Model Blending, Improvements that can be made include Feature Selection, Handling missing data more efficiently, Hyper-parameter tuning, Blending.