Iterative Imputation Using Ratio-based Imputation for High Missing Gap

Intelligent techniques have been designed to learn relying upon complete data. However, sensing error, connection failures, hardware fault, meteorological extremes, etc. lead data to be incomplete, making incomplete data value is a crucial problem in every research domain, including cybersecurity. Incomplete analysis lacks various useful information resulting in poor analysis and estimation. Multiple imputation has been potential solution, which accounts for uncertainty and unbiased results. To enhance the accuracy of the imputed data, this paper proposes a new iterative ratio based imputation (IRBI). Results achieved from the RBI technique is used to update the imputed data values obtained through iterations. Experimental results prove that the IRBI can perform well not only on a high missing amount of data but also on the high missing gap by preserving the data trends and structure.