A Study On Recent Missing Value Imputation Techniques Based On Machine Learning And Deep Learning

One of the biggest challenges associated with real-world applications is that there is a lot of incomplete data, which might lead to inaccurate decisions and misleading analysis overall, thus causing a decrease in performance. Thus, imputing these missing values is crucial in a bid to make the data maintain its integrity and reliability. Even though the methods to address these missing values depend on various factors such as nature of data, the extent of the missingness in data, and the desired results from the imputed values, different methods have been employed to impute these missing values. The conventional methods used for imputation are having some limitations that are more apparent in large real-world datasets. Machine learning(ML) and Deep learning(DL) methods for imputation are being used to solve these problems more efficiently. This paper reviews imputation techniques based on ML and DL methods, along with a comparative study of their merits and demerits.