This publication, "A Detailed Analysis of the KDD CUP 99 Data Set," by Mahbod Tavallaee, Ebrahim Bagheri, Wei Lu, and Ali-A. Ghorbani, addresses the limitations of the KDDCUP'99 data set, which is widely used for evaluating anomaly detection systems. The authors identify two significant issues: a high number of redundant records and an uneven distribution of attack types, which can bias learning algorithms and evaluate systems inaccurately. To address these issues, they propose a new data set, NSL-KDD, which consists of selected records from the original KDD data set, ensuring no redundant or duplicate records and a more balanced distribution of attack types. The NSL-KDD data set is publicly available and aims to provide a more reliable benchmark for researchers to compare different intrusion detection methods. The paper includes a detailed statistical analysis of the KDD data set, experimental results, and a discussion of the proposed solutions.This publication, "A Detailed Analysis of the KDD CUP 99 Data Set," by Mahbod Tavallaee, Ebrahim Bagheri, Wei Lu, and Ali-A. Ghorbani, addresses the limitations of the KDDCUP'99 data set, which is widely used for evaluating anomaly detection systems. The authors identify two significant issues: a high number of redundant records and an uneven distribution of attack types, which can bias learning algorithms and evaluate systems inaccurately. To address these issues, they propose a new data set, NSL-KDD, which consists of selected records from the original KDD data set, ensuring no redundant or duplicate records and a more balanced distribution of attack types. The NSL-KDD data set is publicly available and aims to provide a more reliable benchmark for researchers to compare different intrusion detection methods. The paper includes a detailed statistical analysis of the KDD data set, experimental results, and a discussion of the proposed solutions.