| Sarah's Tech & Research Blog
sehee This paper explores the role of privacy-preserving algorithms in mechanism design, specifically using differential privacy to ensure that participants have limited effect on the outcome of the mechanism and limited incentive to lie.
I describe the characteristics of scoring rules.
Networking Core Knowledge Badge Assessment
I explain the concept and privacy objectives of PPSM for processing sensitive data in electricity or gas markets.
I describe the interaction between data stewards and data analysts regarding query release. I analyse different strategies and explain the considerations.
I present mathematical guarantees for how differential privacy provides truthfulness, limited incentives to lie, and collusion resistance in a game-theoretic context. These properties provide mechanism designers with powerful tools to control the strategic behaviour of agents while preserving privacy.
I explain the Laplace distribution and its mechanism.
I provide basic probability tools that are essential for understanding and proving the effectiveness of differential privacy mechanisms.
I discuss whether differential privacy can be free from harm and its security.
I explain a proposal for post-processing differential privacy, and a theorem for size k.