Soft Computing Techniques in Cyber Defense

Abstract

Artificial intelligence is a technology to make the machines human compatible.Various techniques like-heuristic search (Generate and test, Hill climbing, BFS, DFS, Problem reduction, constraint satisfaction, means-ends analysis etc.), game playing, understanding, planning, NLP, Learning, commonsense, predictions and actions, and expert systems are there to make the system intelligent.This paper primarily focuses on the problem of malicious objects in the cyber space and discusses the usage of different AI techniques to overcome the mentioned problem.Every day the information stored in the computer and the information in transit faces threats due to malicious objects which further leads to a big loss.This paper discusses the possibilities to incorporate the AI techniques to analyze the data and finding and restricting the malicious objects.To make a proactive cyber defense system how the system log can be explored to find the malicious object and how the AI techniques like heuristic search can help in this process is discussed.This paper also discusses about the uncertainty of attacks and gives some views to implement AI techniques such as probability and Bayes" theorem, certainty factors and rule based system, Bayesian networks, fuzzy logic etc.