Efficient Classification with Adaptive KNN

Authors: Puning Zhao, Lifeng Lai11007-11014

AAAI 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Numerical experiments have been conducted to validate our theoretical analysis. In this section, we use numerical experiments to validate our theoretical analysis. In particular, we calculate the convergence rates of the adaptive k NN classifier for some common distributions, and create a log-log plot of the estimated excess risk of the new adaptive k NN classifier versus the sample size. Each point in the curves is averaged over 5,000 trials. The results are shown in Figures 1 and 2.
Researcher Affiliation Academia Puning Zhao and Lifeng Lai University of California Davis {pnzhao, lflai}@ucdavis.edu
Pseudocode Yes Algorithm 1 Adaptive k NN classification algorithm
Open Source Code No The paper does not provide any information about open-source code for the described methodology.
Open Datasets No The paper uses synthetic data generated based on specified distributions (e.g., 'Uniform distribution in [ 5, 5]d', 'standard Gaussian distribution', 'standard Laplace distribution', 'triangular distribution') and regression functions (e.g., 'η(x) = sin(x1)') for its numerical experiments. It does not refer to or provide access to any pre-existing publicly available dataset.
Dataset Splits No The paper does not explicitly provide details about training, validation, or test dataset splits. It mentions 'N' as the sample size and that 'Each point in the curves is averaged over 5,000 trials', but no specific percentages or counts for different splits.
Hardware Specification No The paper does not provide specific hardware details (e.g., CPU, GPU models, memory) used for running its experiments.
Software Dependencies No The paper does not provide specific software dependencies or version numbers (e.g., library names with versions) used for the experiments.
Experiment Setup No The paper describes the distributions and regression functions used for numerical examples (e.g., Uniform, Gaussian, sinusoidal regression function) and states that results are 'averaged over 5,000 trials'. It mentions 'kmax' as the only design parameter of their method but does not specify the value used for the numerical experiments or other specific experimental setup details like hyperparameter values.