Chapter 16 – Security and Privacy in Machine Learning

1. Advanced Encryption Standard (AES): AES is one of the strongest encryption algorithms that protects data. AES accepts different key sizes: 128, 192, or 256 bits.

Triple Data Encryption Standard (DES): Triple DES is an encryption method that uses a 56-bit key to encrypt data blocks.

Blowfish: Blowfish is a symmetric-key encryption technique used as an alternative to the DES encryption algorithm. Blowfish is fast and highly effective for data encryption. It splits data, for example, strings and messages, into blocks of 64 bits and encrypts them individually.

2. We can use a model for inference on encrypted data without the need for decryption.
3. The objective of differential privacy (DP) is to ensure that the removal or addition of individual data points does not affect the outcome of the modeling. For example, by adding random noise to a normal distribution, it tries to make the features of individual...