Reply to thread

Message

<blockquote data-quote="ForgottenSeer 107474" data-source="post: 1076020">[USER=32260]@Andy Ful[/USER]WHHL adds a very strong layer of security on user (writable) folders, while it is still posiible to use a third-party anti-virus. Since all infections start from user land. I was thinking ....Assuming that this Microsoft ISG-Smartscreen cloud backend is fed with the data from the users running Malware Defender, it should be wiser to use an antivirus from another big player in the AV (e.g. Norton/Avira/BullGuard/Avast/AVG) as third-party Antivirus) because the number of endpoints providing infection data would be larger.When I recall correctly a bigger observed population results in a&nbsp; more reliable prediction in statistics. This means that the chance of False Negatives would be smaller by using WHHL with another big player (with a large user base). I realize that the relation population size with prediction reliability is not lineair but parabolic, so it may only gain a few percent or tenths of a percent. But the difficulty of security is to close the last percent (or tenths of percent), so even a small gain is a benefit.Is my assumption correct?</blockquote>

[QUOTE="ForgottenSeer 107474, post: 1076020"] [USER=32260]@Andy Ful[/USER] WHHL adds a very strong layer of security on user (writable) folders, while it is still posiible to use a third-party anti-virus. Since all infections start from user land. I was thinking .... Assuming that this Microsoft ISG-Smartscreen cloud backend is fed with the data from the users running Malware Defender, it should be wiser to use an antivirus from another big player in the AV (e.g. Norton/Avira/BullGuard/Avast/AVG) as third-party Antivirus) because the number of endpoints providing infection data would be larger. When I recall correctly a bigger observed population results in a more reliable prediction in statistics. This means that the chance of False Negatives would be smaller by using WHHL with another big player (with a large user base). I realize that the relation population size with prediction reliability is not lineair but parabolic, so it may only gain a few percent or tenths of a percent. But the difficulty of security is to close the last percent (or tenths of percent), so even a small gain is a benefit. Is my assumption correct? [/QUOTE]

Verification