Variance Reduction in Stochastic Particle-Optimization Sampling

Authors: Jianyi Zhang, Yang Zhao, Changyou Chen

ICML 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Our theoretical results are verified by a number of experiments on both synthetic and real datasets.
Researcher Affiliation Academia 1Duke University 2University at Buffalo, SUNY.
Pseudocode Yes Algorithm 1 SAGA-POS; Algorithm 2 SVRG-POS; Algorithm 3 SVRG-POS+
Open Source Code No The paper does not provide a statement about releasing open-source code or a link to a code repository.
Open Datasets Yes We test the proposed algorithms for Bayesian-logistic-regression (BLR) on four publicly available datasets from the UCI machine learning repository: Australian (690-14), Pima (768-8), Diabetic (1151-20) and Susy (100000-18), where (N d) means a dataset of N data points with dimensionality d.
Dataset Splits No The datasets are split into 80% training data and 20% testing data. There is no explicit mention of a separate validation set split.
Hardware Specification No No specific hardware details such as GPU models, CPU types, or memory specifications were mentioned for running experiments.
Software Dependencies No The paper does not provide specific software dependencies or version numbers (e.g., Python, PyTorch, TensorFlow versions, or library versions).
Experiment Setup Yes Optimized constant stepsizes are applied for each algorithm via grid search. ... The minibatch size is set to 15 for all experiments. ... averaging over 10 runs with 50 particles.