“…Unfortunately, KDD-99 has numerous flaws that make it unsuitable for current applications, including its age, excessively skewed goals, inconstant across training and test data sets, pattern duplication and irrelevant features. Tavallaee et al (2009) proposed NSL-KDD, a more balanced resampling of KDD-99 (Saba, 2020; Begli et al , 2019; Alrawashdeh and Purdy, 2016; Lin et al , 2018; Yang et al , 2019; Schuartz et al , 2020) have considered the KDD-Cup99 data set to demonstrate machine learning and deep learning algorithms to evaluate the proposed method performance, as mentioned in Table 1. But both data set lacks contemporary cyber-attack samples and is considered outdated and does not include recent IoT botnet attack samples.…”