Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1]
Structure from Randomness in Halfspace Learning with the Zero-One Loss
Authors: Ata Kaban, Robert J. Durrant
JAIR 2020 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our main findings are the following: i) In both settings, the obtained bounds are able to discover and take advantage of benign geometric structure, which turns out to depend on the cosine similarities between the classifier and points of the input space, and provide a new interpretation of margin distribution type arguments. ii) Furthermore our bounds allow us to draw new connections between several existing successful classification algorithms, and we also demonstrate that our theory is predictive of empirically observed performance in numerical simulations and experiments. iii) Taken together, these results suggest that the study of compressive learning can improve our understanding of which benign structural traits if they are possessed by the data generator make it easier to learn an effective classifier from a sample. |
| Researcher Affiliation | Academia | Ata Kab an EMAIL School of Computer Science, University of Birmingham Edgbaston, B15 2TT, Birmingham, UK Robert J. Durrant EMAIL Department of Mathematics and Statistics, University of Waikato Hamilton 3240, New Zealand |
| Pseudocode | No | The paper describes algorithms and derivations but does not present any structured pseudocode or algorithm blocks. For example, Section 3.3, 'An Empirical Assessment of Theorem 3.1', details how the algorithm works by describing gradients and optimization, but it's not formatted as pseudocode. |
| Open Source Code | No | We used numerical integration (Simpson quadrature, cf. Mat Lab s built-in function quad ) to evaluate the objective, and a freely available generic nonlinear optimizer4 that employs a combination of conjugate gradient and line search methods. 4. http://learning.eng.cam.ac.uk/carl/code/minimize/ |
| Open Datasets | Yes | Further experimental tests on UCI data sets (Dua & Graff, 2017) are presented in Table 1. |
| Dataset Splits | Yes | For each data set we performed 50 independent splits into two halves. Our parameter k, and SVM s C parameter were set by 5-fold cross-validation on the training half. The error rates on the held-out testing half of the data are reported in Table 1 in comparison with those of SVM (linear kernel). |
| Hardware Specification | No | The paper does not provide any specific details about the hardware used for running the experiments, such as GPU or CPU models, or memory specifications. |
| Software Dependencies | No | The paper mentions "Mat Lab s built-in function quad" and "a freely available generic nonlinear optimizer" but does not specify version numbers for MATLAB or the optimizer. |
| Experiment Setup | Yes | Our parameter k, and SVM s C parameter were set by 5-fold cross-validation on the training half. We initialize z in a low density region (e.g. the mid-point between data centres, z0) and fine-tune it from the data in a small neighbourhood of z0. |