Method and system for detecting malicious behavioral patterns in a computer, using machine learning
Method for detecting malicious behavioral patterns which are related to malicious software such as a computer worm in computerized systems that include data exchange channels with other systems over a data network. Accordingly, hardware and/or software parameters are determined in the computerized system that is can characterize known behavioral patterns thereof. Known malicious code samples are learned by a machine learning process, such as decision trees and artificial neural networks, and the results of the machine learning process are analyzed in respect to the behavioral patterns of the computerized system. Then known and unknown malicious code samples are identified according to the results of the machine learning process.
1. A method for detecting malicious behavioral patterns which are related to malicious software in computerized systems that include data exchange channels with other systems over a data network comprising:
(a) determining a set of parameters comprising at least one hardware parameter and at least one software parameter in said computerized system,
wherein that at least one hardware parameter is selected from a group consisting of processor features and Network Interface features,
wherein the at least one software parameter is selected from a group consisting of TCP layer features, UDP layer features and IP layer features;
wherein the set of parameters is capable of characterizing known behavioral patterns of the computerized system;
(b) performing a first machine learning process and a second machine learning process on the determined set of parameter caused by known malicious code samples, wherein the first machine learning process is provided with input samples of the set of parameters gathered at a first sampling frequency and wherein the second machine learning process is provided with input samples of the set of parameters gathered at a second sampling frequency;
(c) analyzing the results of the first machine learning process and the second machine learning process in respect to the behavioral patterns of said computerized system; and
(d) identifying known and unknown malicious code samples according to the behavioral patterns, which behavioral patters are the results of said first machine learning process and the second machine learning process, wherein the behavioral patterns represent the behavior of the computerized system.
2. The method of claim 1 , wherein the malicious software is a computer worm.