Aligned Objective for Soft-Pseudo-Label Generation in Supervised Learning

Authors: Ning Xu, Yihao Hu, Congyu Qiao, Xin Geng

ICML 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on the benchmark datasets validate the effectiveness of the proposed framework.4. Experiments
Researcher Affiliation Academia 1School of Computer Science and Engineering, Southeast University, Nanjing, China 2Key Laboratory of New Generation Artificial Intelligence Technology and Its Interdisciplinary Applications (Southeast University), Ministry of Education, China. E-mail: {xning, yhhu, qiaocy, xgeng}@seu.edu.cn.
Pseudocode Yes Algorithm 1 SEAL Algorithm
Open Source Code Yes Source code is available at https://github.com/palm-ml/SEAL
Open Datasets Yes We employ three benchmark datasets for multi-class classification including CIFAR-10, CIFAR-100 (Krizhevsky et al., 2009), and Tiny Image Net (Le & Yang, 2015), to evaluate the proposed approach.
Dataset Splits Yes We allocated 10% of the training data from each dataset for validation purposes.
Hardware Specification No The paper does not provide specific hardware details such as GPU or CPU models used for running the experiments.
Software Dependencies No The paper mentions software components like SGD optimizer and backbone networks (ResNet), but does not provide specific version numbers for any software dependencies or libraries.
Experiment Setup Yes We configure the total number of epochs as 200 and set the batch size to 128. We employ the SGD optimizer with a momentum of 0.9 and a weight decay of 1e-4, where the initial learning rate is established at 0.1 with a decay factor of 10%.