CROSS Dataset: Multi-Platform System Logs and Preprocessed Data for Self-Healing Evaluation

This dataset supports the evaluation and development of CROSS (Cross-platform Remediation and Observability Self-Healing System). It includes: Raw logs collected from Android, Linux, macOS, and Windows systems under real and simulated fault conditions. Extracted features in structured CSV format for anomaly classification. Preprocessed data used to train and test the Multinomial Naive Bayes (MNB) classifier embedded in the CROSS framework. The dataset was used for empirical performance analysis and benchmarking of fault detection, resource profiling, and automated remediation success across cyber-physical platforms. For implementation code and metrics, refer to the companion repository: https://github.com/obinnajohnphill/self-healing-remedial-actions