Quasi-Monte Carlo Feature Maps for Shift-Invariant Kernels

Authors: Jiyan Yang, Vikas Sindhwani, Haim Avron, Michael Mahoney

ICML 2014 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Our theoretical analyses are complemented with empirical results that demonstrate the effectiveness of classical and adaptive QMC techniques for this problem. 6. Experiments: In this section we report experiments with both classical QMC sequences and adaptive sequences learnt from box discrepancy minimization.
Researcher Affiliation Collaboration Jiyan Yang1 JIYAN@STANFORD.EDU ICME, Stanford University, Stanford, CA 94305. Vikas Sindhwani1 VSINDHW@US.IBM.COM IBM T. J. Watson Research Center, Yorktown Heights, NY 10598. Haim Avron1 HAIMAV@US.IBM.COM IBM T. J. Watson Research Center, Yorktown Heights, NY 10598. Michael W. Mahoney MMAHONEY@ICSI.BERKELEY.EDU International Computer Science Institute and Dept. of Statistics, University of California at Berkeley, Berkeley, CA 94720
Pseudocode Yes Algorithm 1 Quasi-Random Fourier Features
Open Source Code No The paper mentions using implementations available in MATLAB and publicly available implementations for Lattice Rules and Digital Nets, but it does not provide a link to the authors' own source code for their proposed methodology or experiments.
Open Datasets No The paper names datasets (cpu, census, USPST, mnist) but does not provide specific links, DOIs, repositories, or formal citations for public access to these datasets.
Dataset Splits Yes The ridge parameter is set by the optimal values we obtain via crossvalidation on the MC sequence. The ridge parameter is set by the value which is near-optimal for both sequences in cross-validation.
Hardware Specification No The paper does not provide any specific hardware details such as GPU/CPU models, memory, or processor types used for running its experiments.
Software Dependencies No The paper mentions using 'implementation available in MATLAB' and 'publicly available implementations' for QMC sequences, but it does not specify version numbers for MATLAB or the QMC libraries, or any other software dependencies with versions.
Experiment Setup Yes The ridge parameter is set by the optimal values we obtain via cross-validation on the MC sequence. The Halton sequence is used as the initial setting of the optimization variables S.