Label Distribution Learning by Optimal Transport
Authors: Peng Zhao, Zhi-Hua Zhou
AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental results on real-world datasets comparing with several state-of-the-art methods validate the effectiveness of our approach. |
| Researcher Affiliation | Academia | Peng Zhao, Zhi-Hua Zhou National Key Laboratory for Novel Software Technology, Nanjing University, Nanjing 210023, China Collaborative Innovation Center of Novel Software Technology and Industrialization, Nanjing 210023, China {zhaop, zhouzh}@lamda.nju.edu.cn |
| Pseudocode | Yes | Algorithm 1 Learning the Mapping |
| Open Source Code | No | The paper does not provide any statement about making the source code publicly available or a link to a code repository. |
| Open Datasets | Yes | The 15 datasets cover fields of biological information classification, natural scene recognition, emotional analysis and so on. Due to the page limitation, we only present the brief statistics of the datasets in Table 1. |
| Dataset Splits | Yes | They are chosen by 10-fold cross-validation with random splitting 70% for training and 30% for testing. |
| Hardware Specification | No | The paper does not provide any specific details about the hardware used to run the experiments. |
| Software Dependencies | No | The paper mentions algorithms like 'Sinkhorn-Knopp algorithm' but does not specify any software dependencies with version numbers. |
| Experiment Setup | Yes | Parameter Settings For LALOT, there are two parameters. The first one is the trade-off parameter C, and the other is the entropic regularization coefficient λ. They are chosen by 10-fold cross-validation with random splitting 70% for training and 30% for testing. One more attention is that λ should be chosen starting with a small value. Specifically, λ should small enough to make sure λ M 200 because of the insufficient numerical precision, under which circumstance the Sinkhorns algorithm could blow up due to a too large λ value (Cuturi 2013). |