GoT: a Growing Tree Model for Clustering Ensemble
Authors: Feijiang Li, Yuhua Qian, Jieting Wang8349-8356
AAAI 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We execute two types of experiments to evaluate the performance of Go T. Firstly, we show the working mechanism of Go T on two-dimensional synthetic data sets. Then, we compare Go T with other state-of-art clustering ensemble algorithms on UCI benchmark data sets and image data sets. |
| Researcher Affiliation | Academia | Feijiang Li, Yuhua Qian, Jieting Wang Institute of Big Data Science and Industry, Shanxi University feijiangli@email.sxu.edu.cn, yuhuaqian@126.com, jietingwang@email.sxu.edu.cn |
| Pseudocode | Yes | Algorithm 1 Discovering prototype examples Algorithm 2 Growing Trees Algorithm 3 The Growing Tree Model (Go T) |
| Open Source Code | No | The paper does not provide any explicit statements about making its source code available, nor does it include links to a code repository. |
| Open Datasets | Yes | Eight real data sets from UCI and eight benchmark image data sets are used in this comparison experiment. The detailed information about the eight UCI data sets and the eight image data sets are shown in Table 1. Table 1 includes sources for image data sets such as "Matlab", "UIUC Ponce Research group", "Oxford Visual geometry Group", "(Ciocca et al. 2014)", "Columbia University Image Library", "(Li et al. 2013)", "(Lecun et al. 1998)". |
| Dataset Splits | No | The paper mentions generating "base clustering results" and "ensemble sets" and setting the cluster number and ensemble size, but it does not specify standard training, validation, or test dataset splits (e.g., 80/10/10 percentages or specific sample counts for each split). |
| Hardware Specification | No | The paper does not specify any hardware details (e.g., GPU models, CPU types, or memory) used for running the experiments. |
| Software Dependencies | No | The paper mentions several algorithms and tools (e.g., "k-means algorithm", "Dijkstra’s algorithm", "Otsu’s algorithm", "VGG-16 convolutional neural network", "T-SNE"), but it does not provide specific version numbers for any of these software dependencies. |
| Experiment Setup | Yes | In the experiment, the base clustering results are generated by k-means algorithm with random initial centers. The clusters number of each result is set as min{ n, 50}. The ensemble size is set as l = 50. For each data set, we generate 50 ensemble sets and report the average performance of each compared method. |