Concentration Distribution Learning from Label Distributions
Authors: Jiawei Tang, Yuheng Jia
ICML 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments prove that the proposed approach is able to extract background concentrations from label distributions while producing more accurate prediction results than the stateof-the-art LDL methods. The code is available in https://github.com/seutjw/CDL-LD. ... Extensive experiments on the benchmark datasets clearly show that the LD predicted by our method is better than that predicted by state-of-the-art LDL methods, and the recovery results of our method on background concentrations also fit well with reality. ... 4. Experiments |
| Researcher Affiliation | Academia | 1School of Computer Science and Engineering, Southeast University, Nanjing 210096 2Key Laboratory of New Generation Artificial Intelligence Technology and Its Interdisciplinary Applications (Southeast University), Ministry of Education, China. Correspondence to: Yuheng Jia <EMAIL>. |
| Pseudocode | No | The paper describes the proposed model with mathematical equations and a framework diagram (Figure 3), but it does not include an explicit pseudocode block or algorithm steps formatted like code. |
| Open Source Code | Yes | The code is available in https://github.com/seutjw/CDL-LD. |
| Open Datasets | Yes | The experiments are carried out on 12 real-world datasets with label distribution. The statistics of these datasets are summarized in Table 1. Among these, the first eight (from Alpha to Spoem) are from the clustering analysis of genomewide expression in Yeast Saccharomyces cerevisiae (Eisen et al., 1998). The SJAFFE is collected from JAFFE (Lyons et al., 1998), and the SBU 3DFE is obtained from BU 3DFE (Yin et al., 2006). The Scene consists of multi-label images, where the label distributions are transformed from rankings (Geng & Xia, 2014). |
| Dataset Splits | Yes | We run each method for ten-fold cross-validation. |
| Hardware Specification | Yes | The hardware configuration of the test machine is as follows: AMD EPYC 7K62 48-Cores CPU, 377G running memory and NVIDIA Ge Force RTX 3090 GPU. |
| Software Dependencies | No | The paper does not provide specific software dependencies with version numbers (e.g., Python, PyTorch, TensorFlow, CUDA versions) needed to replicate the experiment. It mentions machine learning paradigms and specific models, but not the underlying software stack. |
| Experiment Setup | Yes | The suggested parameters are used for LDLLC, IIS-LLD, LCLR, LDLFs and DLDL. For LDLLDM, λ1, λ2 and λ3 are tuned from {10-3, 10-2, ..., 103}, and g is tuned from 1 to 14. |