Machine learning in cybersecurity: a comprehensive survey

Abstract

Today’s world is highly network interconnected owing to the pervasiveness of small personal devices (e.g., smartphones) as well as large computing devices or services (e.g., cloud computing or online banking), and thereby each passing minute millions of data bytes are being generated, processed, exchanged, shared, and utilized to yield outcomes in specific applications. Thus, securing the data, machines (devices), and user’s privacy in cyberspace has become an utmost concern for individuals, business organizations, and national governments. In recent years, machine learning (ML) has been widely employed in cybersecurity, for example, intrusion or malware detection and biometric-based user authentication. However, ML algorithms are vulnerable to attacks both in the training and testing phases, which usually leads to remarkable performance decreases and security breaches. Comparatively, limited studies have been conducted to understand the essence and degree of the vulnerabilities of ML techniques against security threats and their defensive mechanisms. It is imperative to systematize recent works related to cybersecurity using ML to seek the attention of researchers, scientists, and engineers. Therefore, in this paper, we provide a comprehensive survey of the works that have been carried out most recently (from 2013 to 2018) on ML in cybersecurity, describing the basics of cyber-attacks and corresponding defenses, the basics of the most commonly used ML algorithms, and proposed ML and data mining schemes for cybersecurity in terms of features, dimensionality reduction, and classification/detection techniques. In this context, this article also provides an overview of adversarial ML, including the security characteristics of deep learning methods. Finally, open issues and challenges in cybersecurity are highlighted and potential future research directions are discussed.

The future of Artificial Intelligence in Cybersecurity: A Comprehensive Survey

“…Dipankar et al [50] included a comprehensive survey of works on cyber security ML (2013 to 2018), the bases and corresponding protection of cyber-attacks, the basics of the most popular ML algorithms and proposed ML and cyber security data mining schemes for features, dimension reduction and classification and detection. This article also offers a summary of adversary ML, including safety features of deep learning methods.…”

Perceived Vulnerability As a Determinant of Increased Risk for Cybersecurity Risk Behavior

“…CS professionals are constantly instituting and upgrading both automatic and user-augmented processes to mitigate risks, such as through patching problematic software, developing new and user-friendly authentication techniques, and machine learning techniques that help identify and minimize the harmful impact of bad actors. [14][15][16] These efforts sometimes go undetected by end users because automatic processes that have been refined over the years tend to be successful at defending against low-level threats that most standard antivirus is capable of neutralizing. 17 Even on an organizational level, a company might implement algorithmically-based policies to minimize automatic downloading of images or files for incoming emails flagged as potential spam, rather than relying on the human to determine how risky clicking on an unsolicited attachment might be.…”

Determining Sequence of Image Processing Technique (IPT) to Detect Adversarial Attacks

“…Some other notable works in that year are meta-learningbased robust detection methods to detect new adversarial attacks with limited examples developed by [44]. Another important and influential work done by [15], where they tried to keep the records of query and used KNN to co-relate that with adversarial examples In summary, defenses against adversarial attacks can be classified in two ways: attack detection and robust recognition model [19]. Detection is a binary classification problem, where input is classified as adversarial or not.…”