A systematic literature review of cyber-security data repositories and performance assessment metrics for semi-supervised learning

Abstract

In Machine Learning, the datasets used to build models are one of the main factors limiting what these models can achieve and how good their predictive performance is. Machine Learning applications for cyber-security or computer security are numerous including cyber threat mitigation and security infrastructure enhancement through pattern recognition, real-time attack detection, and in-depth penetration testing. Therefore, for these applications in particular, the datasets used to build the models must be carefully thought to be representative of real-world data. However, because of the scarcity of labelled data and the cost of manually labelling positive examples, there is a growing corpus of literature utilizing Semi-Supervised Learning with cyber-security data repositories. In this work, we provide a comprehensive overview of publicly available data repositories and datasets used for building computer security or cyber-security systems based on Semi-Supervised Learning, where only a few labels are necessary or available for building strong models. We highlight the strengths and limitations of the data repositories and sets and provide an analysis of the performance assessment metrics used to evaluate the built models. Finally, we discuss open challenges and provide future research directions for using cyber-security datasets and evaluating models built upon them.

U-Net Convolutional Neural Network for Real-Time Prediction of the Number of Cultured Corneal Endothelial Cells for Cellular Therapy

“…The primary motivation for choosing semi-supervised learning is that labeling the data is time-consuming and/or expensive, whereas obtaining unlabeled data requires little time or expense. Semi-supervised learning is widely used in multiple research fields, including the medical field, where it is applied, for example, in oncology diagnostics and care [22,23]. The classification of histopathology and radiotherapy images is essential for various kinds of cancers, such as breast, lung, gastric, liver, colorectal, kidney, pancreatic, and uterine cancers [24][25][26][27][28][29], but adequate labeling by experts is often time-consuming and thus cost-ineffective.…”

Unraveling the Ethical Enigma: Artificial Intelligence in Healthcare

“…SSL is a specific type of ML that requires only a limited number of labels or may work with partially labeled data to build these models. It is crucial that the datasets used in developing these models for cybersecurity accurately represent real-world data to ensure their effectiveness and relevance [ 34 ]. As a fundamental framework for understanding the links and interdependencies among various cybersecurity components from a complete cybersecurity perspective, Grobler et al offered the 3U's of cybersecurity: user, usage, and usability.…”