From: IHVFL: a privacy-enhanced intention-hiding vertical federated learning framework for medical data
Notations | Descriptions |
---|---|
\({\mathcal {C}}\) | Active party in VFL |
\({\mathcal {S}}\) | Passive party in VFL |
\({\mathcal {F}}\) | Intention of active party |
d | Target features |
s | Target samples |
\(\sigma\) | Features statement |
\(\rho\) | Screening vector |
\(\tau\) | Conditional vector |
\(\lambda\) | Random scalar |
I | The data set with target shares |
\(L_1\) | The index set with target samples |
\(L_2\) | The index set with aligned samples |
\(\pi ,\pi ^{-1}\) | Permutations and inverses of permutations |
\(\llbracket x \rrbracket\) | The ciphertext of x |
\(\langle x \rangle\) | The shares of x |