List three synthetic data generation architectures and explain at least one in more detail! $$ Explain the main primitives used for K$-$Anonymity and a few techniques to achieve them when the data is $($non$-)$heterogeneous$!$ Name six anonymization primitives with examples!