Machine Learning in Cyber Trust: Security, Privacy, and Reliability

Many networked computer systems are far too vulnerable to cyber attacks that can inhibit their functioning, corrupt important data, or expose private information. Not surprisingly, the field of cyber-based systems is a fertile ground where many tasks can be formulated as learning problems and approached in terms of machine learning algorithms.

This book contains original materials by leading researchers in the area and covers applications of different machine learning methods in the reliability, security, performance, and privacy issues of cyber space. It enables readers to discover what types of learning methods are at their disposal, summarizing the state-of-the-practice in this significant area, and giving a classification of existing work.

Those working in the field of cyber-based systems, including industrial managers, researchers, engineers, and graduate and senior undergraduate students will find this an indispensable guide in creating systems resistant to and tolerant of cyber attacks.

Many networked computer systems are far too vulnerable to cyber attacks that can inhibit their functioning, corrupt important data, or expose private information. Not surprisingly, the field of cyber-based systems turns out to be a fertile ground where many tasks can be formulated as learning problems and approached in terms of machine learning algorithms.

This book contains original materials by leading researchers in the area and covers applications of different machine learning methods in the security, privacy, and reliability issues of cyber space. It enables readers to discover what types of learning methods are at their disposal, summarizing the state of the practice in this important area, and giving a classification of existing work.

Specific features include the following:

• A survey of various approaches using machine learning/data mining techniques to enhance the traditional security mechanisms of databases
• A discussion of detection of SQL Injection attacks and anomaly detection for defending against insider threats
• An approach to detecting anomalies in a graph-based representation of the data collected during the monitoring of cyber and other infrastructures
• An empirical study of seven online-learning methods on the task of detecting malicious executables
• A novel network intrusion detection framework for mining and detecting sequential intrusion patterns
• A solution for extending the capabilities of existing systems while simultaneously maintaining the stability of the current systems
• An image encryption algorithm based on a chaotic cellular neural network to deal with information security and assurance
• An overview of data privacy research, examining the achievements, challenges and opportunities while pinpointing individual research efforts on the grand map of data privacy protection
• An algorithm based on secure multiparty computation primitives to compute the nearest neighbors of records in horizontally distributed data
• An approach for assessing the reliability of SOA-based systems using AI reasoning techniques
• The models, properties, and applications of context-aware Web services, including an ontology-based context model to enable formal description and acquisition of contextual information pertaining to service requestors and services

Those working in the field of cyber-based systems, including industrial managers, researchers, engineers, and graduate and senior undergraduate students will find this an indispensable guide in creating systems resistant to and tolerant of cyber attacks.