David Jules Froelicher

Biography

I am a research assistant and PhD student under the supervision of Prof. Jean-Pierre Hubaux at the Laboratory for Communications and Applications (LCA1) and Bryan Ford at the Decentralized and Distributed Systems Laboratory (DeDiS), at the Ecole Polytechnique Fédérale de Lausanne (EPFL). I earned my MSc and BSc in Computer Science with a specialisation in IT Security from EPFL in 2016. In 2015, I did a master thesis internship in the NEC research laboratory in Heidelberg, Germany, where I have been involved in the design and implementation of a system enabling proofs of retrievability on deduplicate data. I am currently working on privacy-preserving data sharing by relying on homomorphic encryption, differential privacy, distributed systems and blockchain technologies.

Official source

https://people.epfl.ch/david.froelicher

About this result

This page is automatically generated and may contain information that is not correct, complete, up-to-date, or relevant to your search query. The same applies to every other page on this website. Please make sure to verify the information with EPFL's official sources.

Related units (4)

Related units

Courses taught by this person

Related research domains (10)

Related research domains

Related publications (10)

Related publications

People doing similar research (106)

People doing similar research

, , , ,

Sharing data across multiple institutions for genome-wide association studies (GWAS) would enable discovery of novel genetic variants linked to health and disease. However, existing regulations on genomic data sharing and the sheer size of the data limit the scope of such collaborations. Although cryptographic tools for secure computation promise to enable collaborative studies with formal privacy guarantees, existing approaches either are computationally impractical or support only simplified analysis pipelines. Here, we introduce secure and federated genome-wide association studies (SF-GWAS), a novel combination of secure computation frameworks that empowers efficient and accurate GWAS in a federated manner, i.e., on private data locally-held by multiple entities, while provably ensuring end-to-end data confidentiality. Another key advance is that we designed SF-GWAS to support the two most widely-used GWAS pipelines—those based on principal component analysis (PCA) or linear mixed models (LMMs). We ran SF-GWAS on five real GWAS datasets, including a large UK Biobank cohort of 410K individuals, thereby demonstrating the largest secure genetics collaboration to date. SF-GWAS achieves an order-of-magnitude runtime improvement over the prior art for PCA-based GWAS and newly allows secure LMM-based association tests, for which its runtime scales at a near-constant rate in cohort size. Our work realizes the power of secure, collaborative, and accurate GWAS at unprecedented scale and should be applicable to a broad range of analyses. Our open-source software is at: https://github.com/hhcho/sfgwas.

2022

System and method for privacy-preserving distributed training of neural network models on distributed datasets

Jean-Philippe Léonard Bossuat, David Jules Froelicher, Jean-Pierre Hubaux, Apostolos Pyrgelis, Sinem Sav, Juan Ramón Troncoso-Pastoriza

A computer-implemented method and a distributed computer system (100) for privacy- preserving distributed training of a global neural network model on distributed datasets (DS1 to DSn). The system has a plurality of data providers (DP1 to DPn) being communicatively coupled. Each data provider has a respective local training dataset (DS1 to DSn) and a vector of output labels (OL1 to OLn) for training the global model. Further, it has a portion of a cryptographic distributed secret key (SK1 to SKn) and a corresponding collective cryptographic public key (CPK) of a multiparty fully homomorphic encryption scheme, with the weights of the global model being encrypted with the collective public key. Each data provider (DP1) computes and aggregates, for each layer of the global model, encrypted local gradients (LG1) using the respective local training dataset (DS1) and output labels (OL1), with forward pass and backpropagation using stochastic gradient descent. At least one data provider homomorphically combines at least a subset of the current local gradients of at least a subset of the data providers into combined local gradients, and updates the weights of the current global model (GM) based on the combined local gradients.

2022

POSEIDON: Privacy-Preserving Federated Neural Network Learning

Jean-Philippe Léonard Bossuat, David Jules Froelicher, Joao André Gomes de Sá E Sousa, Jean-Pierre Hubaux, Apostolos Pyrgelis, Sinem Sav, Juan Ramón Troncoso-Pastoriza

In this paper, we address the problem of privacy-preserving training and evaluation of neural networks in an N-party, federated learning setting. We propose a novel system, POSEIDON, the first of its kind in the regime of privacy-preserving neural network training. It employs multiparty lattice-based cryptography to preserve the confidentiality of the training data, the model, and the evaluation data, under a passive-adversary model and collusions between up to N - 1 parties. To efficiently execute the secure backpropagation algorithm for training neural networks, we provide a generic packing approach that enables Single Instruction, Multiple Data (SIMD) operations on encrypted data. We also introduce arbitrary linear transformations within the cryptographic bootstrapping operation, optimizing the costly cryptographic computations over the parties, and we define a constrained optimization problem for choosing the cryptographic parameters. Our experimental results show that POSEIDON achieves accuracy similar to centralized or decentralized non-private approaches and that its computation and communication overhead scales linearly with the number of parties. POSEIDON trains a 3-layer neural network on the MNIST dataset with 784 features and 60K samples distributed among 10 parties in less than 2 hours.

INTERNET SOC2021