Decomposition-Invariant Conditional Gradient for General Polytopes with Line Search

Authors: Mohammad Ali Bashiri, Xinhua Zhang

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

Reproducibility Variable Result LLM Response
Research Type Experimental In this section we compare the empirical performance of AFW-2 against related methods. We first illustrate the performance on kernel binary SVM, then we investigate a problem whose domain is not an SLP, and finally we demonstrate the scalability of AFW-2 on a large scale dataset.
Researcher Affiliation Academia Mohammad Ali Bashiri Xinhua Zhang Department of Computer Science, University of Illinois at Chicago Chicago, Illinois 60607 {mbashi4,zhangx}@uic.edu
Pseudocode Yes Algorithm 1: Decomposition-invariant Away-step Frank-Wolfe (AFW) [...] Algorithm 2: Decomposition-invariant Pairwise Frank-Wolfe (PFW) (exactly the same as [18])
Open Source Code No No explicit statement about providing open-source code or a direct link to a code repository for the described methodology was found.
Open Datasets Yes Three datasets are used. breast-cancer and a1a are obtained from the UCI repository [28] with n = 568 and 1, 605 training examples respectively, and ijcnn1 is from [29] with a subset of 5, 000 examples. [...] M. Lichman. UCI machine learning repository, 2013. URL http://archive.ics.uci.edu/ ml.
Dataset Splits No No explicit details about training/validation/test dataset splits (e.g., percentages, sample counts, or cross-validation setup) are provided. The paper mentions training examples and optimizing test accuracy, but not a distinct validation split.
Hardware Specification No No specific hardware details (e.g., CPU/GPU models, memory amounts, or detailed computer specifications) used for running experiments were mentioned in the paper.
Software Dependencies No No specific software dependencies with version numbers (e.g., library names like PyTorch 1.9, or specific solver versions) were explicitly provided for the experimental setup.
Experiment Setup Yes We first run AFW-2 on the RC-Hull objective in (23), with the value of K set to optimize the test accuracy (K shown in Figure 1). [...] Figure 1: Comparison of SMO and AFW-2 on three different datasets (a) Breast-cancer (K = 10) (b) a1a (K = 30) (c) ijcnn1 (K = 20) [...] Figure 3: Full ijcnn1 (K = 100)