Hands-On Artificial Intelligence for Cybersecurity

Missing values in datasets

One of the most frequent problems that needs to be addressed during the data quality process has to do with missing values ​​within datasets.

This problem occurs, for example, in cases where not all the values ​​of the columns are present, giving rise to null fields. The presence of null fields not only represents a problem for relational databases, but also for many ML algorithms. It is therefore necessary to eliminate these null fields to allow the algorithms to work correctly, without incurring classification errors.

Some of the most common remedies to the problem of missing values ​​include the practice of the following:

• Excluding the lines that have null fields from the dataset
• Excluding the columns that present ...