Academia - Decentriq

Membership Inference as a Model Training Metric

The question of data privacy in machine learning is still widely overlooked. In this blog, we propose a new machine learning training metric that greatly reduces the risk of data reidentification. It can be embedded in any model training with minimal cost.

Academia ·

Closing the circle of security in the cloud

Until today, data-sensitive companies are unable to utilize the immense benefits of cloud computing. Confidential Computing is here to change this. It is well known that cloud-computing offers many advantages, from increased scalability to reduced operations costs.

Academia ·

Proving Knowledge of a Hash Pre-Image with ZoKrates

In this blog we implement a problem very typical for blockchain use-cases: proving the knowledge of a pre-image for a given SHA-256 digest. We will begin demystifying this machinery by computing the SHA-256 hash of the number 5. We will navigate through several options using different languages.

Academia ·

Introduction to zk-SNARKs (Part 1)

In this series of posts, we will look at ZKPs: a family of probabilistic protocols that has garnered increased popularity with the rise of distributed ledger technology (DLT). Let us introduce some of the theory behind this groundbreaking work and the components of its intricate machinery.

Subscribe