We all know that anti-virus software, while certainly important, have their limitations, but the launch of the Israeli security work on algorithms that would give security complexes tougher defenses in the future
All of this revolves around the virus definition updates, security companies regularly deliver their products -. some of them more regularly than others, but there is always a delay in the impact of harmful wild, and antivirus software obtain the corresponding updates to detect it.
According to the MIT Technology Review, in Tel Aviv deep instinct (yes, it sounds more like a movie than a security company) is busy working on their own deep learning algorithms – AI neural networks that are constantly learning as a human the brain, and can recognize malicious fresh as it comes without the need to update the definitions of pipes
Or it certainly ,. Great theory, anyway, though any improvement in the early detection of new types of malicious code will certainly be welcome.
This type of system will be especially helpful when it comes to existing strains of malware that were only slightly modified and tweaked, in an attempt to avoid detection of your typical anti-virus program.
Dr. Eli David, the head of a deep learning research of Deep instinct and the co-founder and chief technology officer, said that in their own tests, the system has already proved that it can detect 20% more new malware than the existing security software.
Of course, other efforts were also made in the deep learning malicious Lamination sphere, in particular, a group of researchers from Microsoft. (Deep learning is a big thing at Redmond, who has a deep technology training Centre working on several initiatives, including a large data analysis, knowledge management and natural language processing).
It may not be too long, then, before we see the technology is refined, and is deployed to significantly strengthen the commercial anti-virus software
Via :. Engadget
- Best antivirus: 10 anti-virus packages and considered rated