Add exponential mechanism and permute-and-flip selection primitives#25
Open
copybara-service[bot] wants to merge 1 commit into
Open
Add exponential mechanism and permute-and-flip selection primitives#25copybara-service[bot] wants to merge 1 commit into
copybara-service[bot] wants to merge 1 commit into
Conversation
Adds two new DPMechanism subclasses for differentially private selection from a score vector: - DPExponentialMechanism: Standard exponential mechanism, preferred under zCDP accounting (tight conversion rho = epsilon^2 / 8). - DPPermuteAndFlip: Permute-and-flip mechanism (McKenna & Sheldon, NeurIPS 2020), which offers expected utility >= exponential mechanism (up to 2x improvement). Under PLD-based accounting, permute-and-flip uses a Laplace privacy loss distribution, giving tighter composition than the exponential mechanism's black-box epsilon-DP bound. Under zCDP, the standard conversion rho = epsilon^2 / 2 applies. Both take a pre-computed score vector (not raw data) and return the index of the selected candidate. PiperOrigin-RevId: 933065255
copybara-service
Bot
force-pushed
the
cl/933065255
branch
from
77f4de8 to
bf21c10
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Add exponential mechanism and permute-and-flip selection primitives
Adds two new DPMechanism subclasses for differentially private selection
from a score vector:
zCDP accounting (tight conversion rho = epsilon^2 / 8).
NeurIPS 2020), which offers expected utility >= exponential mechanism
(up to 2x improvement). Under PLD-based accounting, permute-and-flip
uses a Laplace privacy loss distribution, giving tighter composition
than the exponential mechanism's black-box epsilon-DP bound. Under
zCDP, the standard conversion rho = epsilon^2 / 2 applies.
Both take a pre-computed score vector (not raw data) and return the index
of the selected candidate.