More Like this

1. ELSA: Secure Aggregation for Federated Learning with Malicious Actors

   Rathee, Mayank ; Shen, Conghao ; Wagh, Sameer ; Popa, Raluca Ada ( January 2023 , Proceedings of the IEEE Symposium on Security and Privacy)

   Federated learning (FL) is an increasingly popular approach for machine learning (ML) in cases where the training dataset is highly distributed. Clients perform local training on their datasets and the updates are then aggregated into the global model. Existing protocols for aggregation are either inefficient, or don’t consider the case of malicious actors in the system. This is a major barrier in making FL an ideal solution for privacy-sensitive ML applications. We present ELSA, a secure aggregation protocol for FL, which breaks this barrier - it is efficient and addresses the existence of malicious actors at the core of its design. Similar to prior work on Prio and Prio+, ELSA provides a novel secure aggregation protocol built out of distributed trust across two servers that keeps individual client updates private as long as one server is honest, defends against malicious clients, and is efficient end-to-end. Compared to prior works, the distinguishing theme in ELSA is that instead of the servers generating cryptographic correlations interactively, the clients act as untrusted dealers of these correlations without compromising the protocol’s security. This leads to a much faster protocol while also achieving stronger security at that efficiency compared to prior work. We introduce new techniques that retain privacy even when a server is malicious at a small added cost of 7-25% in runtime with negligible increase in communication over the case of semi-honest server. Our work improves end-to-end runtime over prior work with similar security guarantees by big margins - single-aggregator RoFL by up to 305x (for the models we consider), and distributed trust Prio by up to 8x. 

   more » « less
2. Federated Boosted Decision Trees with Differential Privacy

   https://doi.org/10.1145/3548606.3560687  

   Maddock, Samuel ; Cormode, Graham ; Wang, Tianhao ; Maple, Carsten ; Jha, Somesh ( November 2022 , ACM)

   There is great demand for scalable, secure, and efficient privacy-preserving machine learning models that can be trained over distributed data. While deep learning models typically achieve the best results in a centralized non-secure setting, different models can excel when privacy and communication constraints are imposed. Instead, tree-based approaches such as XGBoost have attracted much attention for their high performance and ease of use; in particular, they often achieve state-of-the-art results on tabular data. Consequently, several recent works have focused on translating Gradient Boosted Decision Tree (GBDT) models like XGBoost into federated settings, via cryptographic mechanisms such as Homomorphic Encryption (HE) and Secure Multi-Party Computation (MPC). However, these do not always provide formal privacy guarantees, or consider the full range of hyperparameters and implementation settings. In this work, we implement the GBDT model under Differential Privacy (DP). We propose a general framework that captures and extends existing approaches for differentially private decision trees. Our framework of methods is tailored to the federated setting, and we show that with a careful choice of techniques it is possible to achieve very high utility while maintaining strong levels of privacy. 

   more » « less
3. ELSA: Secure Aggregation for Federated Learning with Malicious Actors

   Rathee, Mayank ; Shen, Conghao ; Wagh, Sameer ; Popa, Raluca Ada ( January 2023 , Proceedings of the IEEE Symposium on Security and Privacy)

   Federated learning (FL) is an increasingly popular approach for machine learning (ML) when the training dataset is highly distributed. Clients perform local training on their datasets and the updates are then aggregated into the global model. Existing protocols for aggregation are either inefficient or don’t consider the case of malicious actors in the system. This is a major barrier to making FL an ideal solution for privacy-sensitive ML applications. In this talk, I will present ELSA, a secure aggregation protocol for FL that breaks this barrier - it is efficient and addresses the existence of malicious actors (clients + servers) at the core of its design. Similar to prior work Prio and Prio+, ELSA provides a novel secure aggregation protocol built out of distributed trust across two servers that keeps individual client updates private as long as one server is honest, defends against malicious clients, and is efficient end-to-end. Compared to prior works, the distinguishing theme in ELSA is that instead of the servers generating cryptographic correlations interactively, the clients act as untrusted dealers of these correlations without compromising the protocol’s security. This leads to a much faster protocol while also achieving stronger security at that efficiency compared to prior work. We introduce new techniques that retain privacy even when a server is malicious at a small added cost of 7-25% in runtime with a negligible increase in communication over the case of a semi-honest server. ELSA improves end-to-end runtime over prior work with similar security guarantees by big margins - single-aggregator RoFL by up to 305x (for the models we consider), and distributed-trust Prio by up to 8x (with up to 16x faster server-side protocol). Additionally, ELSA can be run in a bandwidth-saver mode for clients who are geographically bandwidth-constrained - an important property that is missing from prior works. 

   more » « less
4. A Hybrid Approach to Secure Function Evaluation using SGX

   https://doi.org/10.1145/3321705.3329835  

   Choi, Joseph I. ; Tian, Dave ; Hernandez, Grant ; Patton, Christopher ; Mood, Benjamin ; Shrimpton, Thomas ; Butler, Kevin R. ; Traynor, Patrick ( July 2019 , ACM Asia Conference on Computer and Communications Security)

   A protocol for two-party secure function evaluation (2P-SFE) aims to allow the parties to learn the output of function f of their private inputs, while leaking nothing more. In a sense, such a protocol realizes a trusted oracle that computes f and returns the result to both parties. There have been tremendous strides in efficiency over the past ten years, yet 2P-SFE protocols remain impractical for most real-time, online computations, particularly on modestly provisioned devices. Intel's Software Guard Extensions (SGX) provides hardware-protected execution environments, called enclaves, that may be viewed as trusted computation oracles. While SGX provides native CPU speed for secure computation, previous side-channel and micro-architecture attacks have demonstrated how security guarantees of enclaves can be compromised. In this paper, we explore a balanced approach to 2P-SFE on SGX-enabled processors by constructing a protocol for evaluating f relative to a partitioning of f. This approach alleviates the burden of trust on the enclave by allowing the protocol designer to choose which components should be evaluated within the enclave, and which via standard cryptographic techniques. We describe SGX-enabled SFE protocols (modeling the enclave as an oracle), and formalize the strongest-possible notion of 2P-SFE for our setting. We prove our protocol meets this notion when properly realized. We implement the protocol and apply it to two practical problems: privacy-preserving queries to a database, and a version of Dijkstra's algorithm for privacy-preserving navigation. Our evaluation shows that our SGX-enabled SFE scheme enjoys a 38x increase in performance over garbled-circuit-based SFE. Finally, we justify modeling of the enclave as an oracle by implementing protections against known side-channels. 

   more » « less
5. Issues in the Reproducibility of Deep Learning Results

   https://doi.org/10.1109/SPMB47826.2019.9037840  

   Jean-Paul, S. ; Elseify, T. ; Obeid, I. ; Picone, J. ( December 2019 , IEEE Signal Processing in Medicine and Biology Symposium (SPMB))

   Obeid, I. ; Selesnik, I. ; Picone, J. (Ed.)

   The Neuronix high-performance computing cluster allows us to conduct extensive machine learning experiments on big data [1]. This heterogeneous cluster uses innovative scheduling technology, Slurm [2], that manages a network of CPUs and graphics processing units (GPUs). The GPU farm consists of a variety of processors ranging from low-end consumer grade devices such as the Nvidia GTX 970 to higher-end devices such as the GeForce RTX 2080. These GPUs are essential to our research since they allow extremely compute-intensive deep learning tasks to be executed on massive data resources such as the TUH EEG Corpus [2]. We use TensorFlow [3] as the core machine learning library for our deep learning systems, and routinely employ multiple GPUs to accelerate the training process. Reproducible results are essential to machine learning research. Reproducibility in this context means the ability to replicate an existing experiment – performance metrics such as error rates should be identical and floating-point calculations should match closely. Three examples of ways we typically expect an experiment to be replicable are: (1) The same job run on the same processor should produce the same results each time it is run. (2) A job run on a CPU and GPU should produce identical results. (3) A job should produce comparable results if the data is presented in a different order. System optimization requires an ability to directly compare error rates for algorithms evaluated under comparable operating conditions. However, it is a difficult task to exactly reproduce the results for large, complex deep learning systems that often require more than a trillion calculations per experiment [5]. This is a fairly well-known issue and one we will explore in this poster. Researchers must be able to replicate results on a specific data set to establish the integrity of an implementation. They can then use that implementation as a baseline for comparison purposes. A lack of reproducibility makes it very difficult to debug algorithms and validate changes to the system. Equally important, since many results in deep learning research are dependent on the order in which the system is exposed to the data, the specific processors used, and even the order in which those processors are accessed, it becomes a challenging problem to compare two algorithms since each system must be individually optimized for a specific data set or processor. This is extremely time-consuming for algorithm research in which a single run often taxes a computing environment to its limits. Well-known techniques such as cross-validation [5,6] can be used to mitigate these effects, but this is also computationally expensive. These issues are further compounded by the fact that most deep learning algorithms are susceptible to the way computational noise propagates through the system. GPUs are particularly notorious for this because, in a clustered environment, it becomes more difficult to control which processors are used at various points in time. Another equally frustrating issue is that upgrades to the deep learning package, such as the transition from TensorFlow v1.9 to v1.13, can also result in large fluctuations in error rates when re-running the same experiment. Since TensorFlow is constantly updating functions to support GPU use, maintaining an historical archive of experimental results that can be used to calibrate algorithm research is quite a challenge. This makes it very difficult to optimize the system or select the best configurations. The overall impact of all of these issues described above is significant as error rates can fluctuate by as much as 25% due to these types of computational issues. Cross-validation is one technique used to mitigate this, but that is expensive since you need to do multiple runs over the data, which further taxes a computing infrastructure already running at max capacity. GPUs are preferred when training a large network since these systems train at least two orders of magnitude faster than CPUs [7]. Large-scale experiments are simply not feasible without using GPUs. However, there is a tradeoff to gain this performance. Since all our GPUs use the NVIDIA CUDA® Deep Neural Network library (cuDNN) [8], a GPU-accelerated library of primitives for deep neural networks, it adds an element of randomness into the experiment. When a GPU is used to train a network in TensorFlow, it automatically searches for a cuDNN implementation. NVIDIA’s cuDNN implementation provides algorithms that increase the performance and help the model train quicker, but they are non-deterministic algorithms [9,10]. Since our networks have many complex layers, there is no easy way to avoid this randomness. Instead of comparing each epoch, we compare the average performance of the experiment because it gives us a hint of how our model is performing per experiment, and if the changes we make are efficient. In this poster, we will discuss a variety of issues related to reproducibility and introduce ways we mitigate these effects. For example, TensorFlow uses a random number generator (RNG) which is not seeded by default. TensorFlow determines the initialization point and how certain functions execute using the RNG. The solution for this is seeding all the necessary components before training the model. This forces TensorFlow to use the same initialization point and sets how certain layers work (e.g., dropout layers). However, seeding all the RNGs will not guarantee a controlled experiment. Other variables can affect the outcome of the experiment such as training using GPUs, allowing multi-threading on CPUs, using certain layers, etc. To mitigate our problems with reproducibility, we first make sure that the data is processed in the same order during training. Therefore, we save the data from the last experiment and to make sure the newer experiment follows the same order. If we allow the data to be shuffled, it can affect the performance due to how the model was exposed to the data. We also specify the float data type to be 32-bit since Python defaults to 64-bit. We try to avoid using 64-bit precision because the numbers produced by a GPU can vary significantly depending on the GPU architecture [11-13]. Controlling precision somewhat reduces differences due to computational noise even though technically it increases the amount of computational noise. We are currently developing more advanced techniques for preserving the efficiency of our training process while also maintaining the ability to reproduce models. In our poster presentation we will demonstrate these issues using some novel visualization tools, present several examples of the extent to which these issues influence research results on electroencephalography (EEG) and digital pathology experiments and introduce new ways to manage such computational issues. 

   more » « less