Machine Learning with Adversaries: Byzantine Tolerant Gradient Descent

Authors: Peva Blanchard, El Mahdi El Mhamdi, Rachid Guerraoui, Julien Stainer

NeurIPS 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We also report on experimental evaluations of Krum.
Researcher Affiliation Academia Peva Blanchard EPFL, Switzerland peva.blanchard@epfl.ch El Mahdi El Mhamdi EPFL, Switzerland elmahdi.elmhamdi@epfl.ch Rachid Guerraoui EPFL, Switzerland rachid.guerraoui@epfl.ch Julien Stainer EPFL, Switzerland julien.stainer@epfl.ch
Pseudocode No The paper describes the Krum function in a prose paragraph but does not present it as a formally labeled pseudocode block or algorithm block formatted like code.
Open Source Code No The implementation is part of a larger distributed framework to run sgd in a reliable distributed fashion and will be released in the github repository of the distributed computing group at epfl, https://github.com/lpd-epfl.
Open Datasets Yes We consider the task of spam filtering (dataset spambase [19]). ... image classification (dataset MNIST). ... [19] M. Lichman. UCI machine learning repository, 2013.
Dataset Splits No We measure the error using cross-validation. The paper mentions cross-validation as an evaluation method, but it does not specify exact split percentages (e.g., 80/10/10), absolute sample counts for validation, or references to predefined validation splits.
Hardware Specification No The paper does not provide specific hardware details such as GPU/CPU models, memory specifications, or cloud computing instance types used for running the experiments. It only refers to a 'distributed machine learning framework'.
Software Dependencies No The paper does not provide specific software dependencies with version numbers (e.g., Python 3.8, PyTorch 1.9, CUDA 11.1) needed to replicate the experiment.
Experiment Setup Yes The learning model is a multi-layer perceptron (MLP) with two hidden layers. ... Each (correct) worker estimates the gradient on a mini-batch of size 3.