Orthogonal Random Features

Authors: Felix Xinnan X. Yu, Ananda Theertha Suresh, Krzysztof M. Choromanski, Daniel N. Holtmann-Rice, Sanjiv Kumar

NeurIPS 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on several datasets verify the effectiveness of ORF and SORF over the existing methods.
Researcher Affiliation Industry Google Research, New York {felixyu, theertha, kchoro, dhr, sanjivk}@google.com
Pseudocode No The paper does not contain any structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide any explicit statement or link regarding the release of open-source code for the described methodology.
Open Datasets Yes We first show kernel approximation performance on six datasets. The input feature dimension d is set to be power of 2 by padding zeros or subsampling. Figure 4 compares the mean squared error (MSE) of all methods. For fixed D, the kernel approximation MSE exhibits the following ordering: SORF ' ORF < QMC [25] < RFF [19] < Other fast kernel approximations [13, 28]. We also apply ORF and SORF on classification tasks. Table 2 shows classification accuracy for different kernel approximation techniques with a (linear) SVM classifier. Datasets mentioned include LETTER, FOREST, USPS, CIFAR, MNIST, GISETTE.
Dataset Splits No The paper mentions using datasets for classification tasks but does not provide specific details on training, validation, or test set splits, nor does it specify cross-validation methods.
Hardware Specification No The paper does not provide any specific details about the hardware (e.g., GPU/CPU models, memory specifications) used for running its experiments.
Software Dependencies No The paper does not specify any software dependencies with version numbers (e.g., programming languages, libraries, or frameworks) required to replicate the experiments.
Experiment Setup Yes For each dataset, σ is chosen to be the mean distance of the 50th 2 nearest neighbor for 1,000 sampled datapoints. Empirically, this yields good classification results. The role of σ: Note that a very small σ will lead to overfitting, and a very large σ provides no discriminative power for classification. Throughout the experiments, σ for each dataset is chosen to be the mean distance of the 50th 2 nearest neighbor, which empirically yields good classification results [28].