Processed IDS Datasets for Reproducible Multiclass Active Learning Experiments

This repository provides processed versions of publicly available intrusion detection datasets prepared for reproducible experiments in multiclass and imbalanced Active Learning scenarios. The datasets follow a standardized preprocessing and partitioning pipeline, as described in the paper “Active Learning for Multiclass Imbalanced Intrusion Detection: Multi-Dataset Evaluation”. All data preprocessing and dataset treatment procedures are publicly available at \url{https://github.com/rodolfoarruda/share/tree/main/datasets_treatment}. The original datasets are properly cited in the associated publication and remain available from their official sources.