On Optimal Generalizability in Parametric Learning

Authors: Ahmad Beirami, Meisam Razaviyayn, Shahin Shahrampour, Vahid Tarokh

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

Reproducibility Variable Result LLM Response
Research Type Experimental In our numerical experiments, we illustrate the accuracy and efficiency of ALOOCV as well as our proposed framework for the optimization of the regularizer.
Researcher Affiliation Academia School of Engineering and Applied Sciences, Harvard University, Cambridge, MA 02138, USA. Department of Industrial and Systems Engineering, University of Southern California, Los Angeles, CA 90089, USA.
Pseudocode Yes Algorithm 1 Approximate gradient descent algorithm for tuning λ; Algorithm 2 Stochastic (online) approximate gradient descent algorithm for tuning λ
Open Source Code No The paper does not provide any specific links to source code repositories or explicitly state that the code is publicly available.
Open Datasets Yes We applied logistic regression on MNIST and CIFAR-10 image datasets
Dataset Splits Yes A classical cross validation strategy is the leave-one-out cross validation (LOOCV) where one sample is left out for validation and training is done on the rest of the samples that are presented to the learner, and this process is repeated on all of the samples.
Hardware Specification No The paper only mentions 'on a PC' without specifying any details about the CPU, GPU, or memory.
Software Dependencies No The paper does not list any specific software dependencies with version numbers.
Experiment Setup Yes We initialize the algorithm with λ1 = ... = λ1 50 = 1/3 and compute ACV using Theorem 1. ... Initialize the tuning parameter λ0, choose a step-size selection rule, and set t = 0.