Pareto Ensemble Pruning
Authors: Chao Qian, Yang Yu, Zhi-Hua Zhou
AAAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We conducted experiments on 20 binary and 10 multiclass data sets (Blake, Keogh, and Merz 1998), pruning the base classifiers trained by Bagging (Breiman 1996). To assess each method on each data set, we repeat the following process 30 times. The data set is randomly and evenly split into three parts, each as the training set, the validation set and the test set. |
| Researcher Affiliation | Academia | Chao Qian and Yang Yu and Zhi-Hua Zhou National Key Laboratory for Novel Software Technology, Nanjing University Collaborative Innovation Center of Novel Software Technology and Industrialization Nanjing 210023, China {qianc,yuy,zhouzh}@lamda.nju.edu.cn |
| Pseudocode | Yes | Algorithm 1 (PEP). Given a set of trained classifiers H = {hi}n i=1, an objective f : 2H R and an evaluation criterion eval, it contains: 1. Let g(s) = (f(Hs), |s|) be the bi-objective. ... Algorithm 2 (VDS Subroutine). Given a pseudo-Boolean function f and a solution s, it contains: 1. Q = , L = . ... |
| Open Source Code | No | The paper does not provide an explicit statement or link for open-source code for the methodology described. |
| Open Datasets | Yes | We conducted experiments on 20 binary and 10 multiclass data sets (Blake, Keogh, and Merz 1998), pruning the base classifiers trained by Bagging (Breiman 1996). ... We employ a lately available MHAR data set, published in (Anguita et al. 2012). |
| Dataset Splits | Yes | The data set is randomly and evenly split into three parts, each as the training set, the validation set and the test set. ... In each run, we fix the test set and randomly split the training set into two parts: 75% as the training and 25% as the validation. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., CPU/GPU models, memory, or specific computing infrastructure) used for running its experiments. |
| Software Dependencies | No | The paper mentions algorithms like 'C4.5 decision trees (Quinlan 1993)' but does not list specific software dependencies with version numbers (e.g., libraries, frameworks, or operating systems). |
| Experiment Setup | Yes | A Bagging of 100 C4.5 decision trees (Quinlan 1993) is trained on the training set... The number of iterations for PEP is set to be n2 log n... The parameter p for MD is set to be 0.075... and the parameter ρ of DREP is selected from {0.2, 0.25, . . . , 0.5}. |