Machine Learning with Swift

Privacy and differential privacy

Surprisingly, in the last few years, most scientific papers where mobile devices and machine learning were mentioned together were not about computer vision or natural language processing. The topics discussed the most were information security and privacy. These two fields intersect in several scenarios:

• The attacker employs offensive machine learning as a part of his/her toolkit. It can be used for discovery and analysis of vulnerabilities or for the attack itself. Examples are face or voice recognition for surveillance and finding data leaks in an improperly anonymized data.
• Defensive machine learning is used to protect against cyber attacks. It can be utilized for both threat detection and analysis. An ...