Firewall log classification is important to monitor network traffic. Machine learning classifiers using network-related features have shown good results for this task, but high-dimensional features slow down classification. In this paper, we apply feature selection using bee swarm optimization with Q-learning-based local search (QBSO-FS) to the UCI Internet Firewall Data Set, together with random forest (RF), k-nearest neighbor (KNN), and Bernoulli naïve Bayes (Bernoulli NB) classifiers. The first version of this preprint selected and evaluated feature subsets on the same 5-fold cross-validation of the full data set, which contains 12.8% exact duplicates; the reported scores are therefore optimistically biased. In this extended version, we re-evaluate the method with a leakage-free protocol: the data are split into training, validation, and test sets, features are selected using the validation set only, and accuracy and macro-averaged scores are measured once on the untouched test set, for both the raw data and a deduplicated, preprocessed version. On a single stratified split, QBSO-FS selects 5--7 features. The selected subsets clearly improve Macro-F1 over all features for KNN on both data sets and for RF on the deduplicated data (by 8--19 percentage points). For RF on the raw data and for Bernoulli NB, the differences (0.2--1.5 percentage points) may not hold for a different split. RF with 7 selected features reaches 99.90% accuracy and 94.88% Macro-F1 on the deduplicated data. On the same split, this exceeds our re-implementations of the deep neural network configurations of previous studies and the feature subset selected by particle swarm optimization in one of them, while the run time saved on this low-dimensional data set is small. These results suggest that the method can support automatic firewall log analysis, while evaluation over multiple splits remains future work.
Background: Since 1990 many feature selection methods have been proposed across heterogeneous applications. To validate the usefulness of a new method, it needs to be compared against at least one baseline method from the existing literature on a feature selection task using at least one dataset. Recent developments in tabular Deep Learning (DL) and data valuation in Machine Learning (ML) suggest that the evaluation of new methods, algorithms, and models may be consciously or unconsciously biased. We hypothesise that a similar trend exists in feature selection (FS), particularly in filter feature selection (FFS). The aim of this study is therefore to examine FFS studies to identify factors that influence the evaluation and that might consist entry point for biases in order to recommend stronger principles for FFS evaluation. Methods: We analyse a sample of 28 high profile FFS studies published between 1994 and 2025. The analysis provides reflections on how to examine FFS studies, highlights lessons learned throughout the process, and gives five evidence-based recommendations for future FFS evaluation. Results: Multivariate Linear Regression analysis achieved a score of R2=0.33. It means that 33% of the variance in the performance of new methods against chosen baselines (win rate) is explained by the number of datasets (#Datasets), the number of baselines (#Baselines), and the number of new methods (#NewMethods). Discussion: R2=0.33 is considered medium explanation; which is promising given that this is the first such study. The medium explanation result is due to the fact that win rate is influenced by additional factors such as the maturity of the feature selection domain, the type of datasets and baselines, and the simplicity of the regression model used to explain the relationship.
Malick Ebiele, Malika Bendechache, Rob Brennan
1*Computer Science, University College Dublin, Dublin, Dublin, Ireland. · 3ADAPT Centre, Dublin, Dublin, Ireland. · 2Computer Science, University of Galway, Galway, Galway, Ireland.
Feature selection is critical for network intrusion detection systems (NIDS) operating under high-dimensional, highly imbalanced traffic, as found in operational and defense networks. Traditional filter methods rank features using global statistics computed symmetrically across classes and thus fail to capture the asymmetry of intrusion detection, where attacks are best characterized as deviations from dominant benign traffic. We propose benign-anchored Classwise Mean Deviation (nCMD), a lightweight and interpretable method that scores feature relevance based on the deviation of attack-class distributions from the benign-class mean, rather than a globally biased reference. This approach aligns feature selection with the operational semantics of NIDS at no additional computational cost. Across four benchmark datasets (CICIDS2017, CICDDoS2019, NSL-KDD, and UNSW-NB15), multiple feature budgets, and three downstream classifiers, nCMD matches or exceeds classical filter baselines in macro-averaged F1-score. It achieves the best result on three of the four datasets and under every classifier, with the strongest improvements observed under tight feature budgets and severe class imbalance. These results support benign-anchored ranking as a scalable and interpretable preprocessing component for resource-constrained NIDS.
Abu Fuad Ahmad, Istiaque Ahmed
Department of Computer Science · New Mexico State University · Las Cruces, New Mexico, USA +3
Weighted sums of random kitchen sinks (W-RKS) is a commonly used feature extraction technique for supervised learning tasks. A major drawback of this method is the redundancy and ineffectiveness of many extracted random features. This paper investigates efficient methods for reducing such redundancy in multi-class classification problems. We specifically formulate feature selection as a variant of the sequentially nested model selection problem and introduce the Successive Empirical Excess Risk Difference Maximization (SEED-MAX) algorithm. It recursively computes empirical risks to identify the most parsimonious yet informative set of features. We derive an analytical lower bound for the probability of correct feature selection by the SEED-MAX algorithm. This lower bound on the probability of correct feature selection can be made arbitrarily close to one by increasing the training dataset size. To further enhance the rate of pruning, we evaluate the features retained by SEED-MAX using the Fisher test (F-Test). Empirical evaluations on UCR time-series classification datasets show that the sequential application of SEED-MAX with the Fisher test achieves approximately 92 percent rate of pruning with negligible degradation in classification accuracy. Implementations of the SEED-MAX and SEED-MAX-F-Test methods are available at https://github.com/OmidiAmin/seed-max-ftest.
Mohammad Amin Omidi, Babak Seyfe, Shahrokh Valaee
Information Theoretic Learning Systems Laboratory (ITLSL), Department of Electrical Engineering, Shahed University, Tehran, Iran · Department of Electrical and Computer Engineering, University of Toronto, Toronto, Canada