Microsoft wants AI to detect malware attacks on Windows PCs before they happen

Microsoft's anti-malware tool, Windows Defender, is often touted by the company as Windows' primary line of defense against threats. The firm continuously updates it to make it more secure, and often cites its methodologies for this endeavor as well.

Now, it appears that Microsoft is looking to utilize the powers of artificial intelligence to counter malware before it infects a Windows PC.

Microsoft is hosting a competition on popular data science website, Kaggle, where it is "challenging the data science community to develop techniques to predict if a machine will soon be hit with malware."

Participants will be provided with 9.4GB of anonymous data from 16.8 million real-world machines as a training set. Using this enormous dataset, data scientists will be tasked with developing a model that achieves the highest accuracy on the test data. $25,000 will be shared among the top five teams in the following way:

  • 1st Place - $12,000
  • 2nd Place - $7,000
  • 3rd Place - $3,000
  • 4th Place - $2,000
  • 5th Place - $1,000

As of right now, the top entry has achieved an accuracy rate of 68.9%, though this will likely improve before the competition closes on March 13, 2019.

It is important to note that this isn't the first competition of its type hosted by Microsoft on Kaggle. Back in 2015, it offered the public a chance to win $16,000 in prizes using 0.5TB of training data in a malware classification challenge. Interestingly, the current contest is hosted by the Windows Defender ATP Research team in particular, which means that the findings of this competition will likely be used to improve Microsoft's own line of defense against malicious threats.

Report a problem with article
1511054096_img_2949
Next Article

Apple seeks to address imminent iPhone sales ban in China with software updates

microsoft-sharepoint
Previous Article

Microsoft SharePoint for iOS gets support for @mention and more

17 Comments - Add comment

Advertisement