ReLISH: Reliable Label Inference via Smoothness Hypothesis
Authors: Chen Gong, Dacheng Tao, Keren Fu, Jie Yang
AAAI 2014 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We conduct comprehensive experiments to compare Re LISH with representative SSL algorithms on several public datasets, including UCI (Frank and Asuncion 2010), the Optical Recognition of Handwritten Digits Dataset (Frank and Asuncion 2010), and Caltech 256 (Griffin, Holub, and Perona 2007). These empirical studies complement our theoretical studies and show that Re LISH achieves promising performance on both the transductive and inductive settings. |
| Researcher Affiliation | Academia | Chen Gong ,? and Dacheng Tao? and Keren Fu and Jie Yang Institute of Image Processing and Pattern Recognition, Shanghai Jiao Tong University ?Centre for Quantum Computation and Intelligent Systems, University of Technology Sydney {goodgongchen, fkrsuper, jieyang}@sjtu.edu.cn dacheng.tao@uts.edu.au |
| Pseudocode | No | The paper does not contain structured pseudocode or algorithm blocks. |
| Open Source Code | No | No explicit statement or link providing access to the open-source code for the described methodology. |
| Open Datasets | Yes | We conduct comprehensive experiments to compare Re LISH with representative SSL algorithms on several public datasets, including UCI (Frank and Asuncion 2010), the Optical Recognition of Handwritten Digits Dataset (Frank and Asuncion 2010), and Caltech 256 (Griffin, Holub, and Perona 2007). |
| Dataset Splits | Yes | We conducted the simulations by using n = 60 training examples. |
| Hardware Specification | No | No specific hardware details (e.g., CPU/GPU models, memory, or cluster specifications) used for running experiments are provided in the paper. |
| Software Dependencies | No | No specific software dependencies with version numbers are mentioned. The paper only implicitly refers to software through algorithm names. |
| Experiment Setup | Yes | We built k-NN graphs with σ empirically tuned to optimal for all the algorithms throughout this section, and the model parameters , β, γ of Re LISH were also properly tuned for each dataset. We also empirically show that the Re LISH performs robustly for a wide range of each of the model parameters in the supplementary material. |