Learning a Gradient-free Riemannian Optimizer on Tangent Spaces

Authors: Xiaomeng Fan, Zhi Gao, Yuwei Wu, Yunde Jia, Mehrtash Harandi7377-7384

AAAI 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We empirically show that our method brings efficient learning of the optimizer, while enjoying a good optimization trajectory in a data-driven manner.In this section, we evaluate our method from three perspectives, that is, the convergence, accuracy and efficiency.
Researcher Affiliation Academia 1 Beijing Laboratory of Intelligent Information Technology School of Computer Science, Beijing Institute of Technology, Beijing, China 2 Department of Electrical and Computer Systems Eng., Monash University, and Data61, Australia
Pseudocode No The paper does not include formal pseudocode or algorithm blocks. Figure 2 and Figure 3 depict architectural diagrams and computational graphs, not step-by-step algorithms.
Open Source Code Yes The code is available at https://github.com/Xiaomeng Fan-mcislab/Learning-a-Gradient-free-Riemannian-Optimizer-on-Tangent-Spaces
Open Datasets Yes We utilize the MNIST dataset to evaluate our optimizer.We use the Yale B dataset (Lee, Ho, and Kriegman 2005) to evaluate our optimizer.We evaluate our optimizer on the Kylberg texture dataset (Kylberg 2011).
Dataset Splits No The paper mentions training and testing but does not explicitly detail a separate validation split or how hyperparameters were tuned using a validation set. It states "hyperparameters of all optimizers are tuned to achieve the best performance" which implies validation, but no specifics are given.
Hardware Specification No The paper does not provide specific details about the hardware used to run the experiments, such as GPU models, CPU types, or memory specifications. It only mentions training times.
Software Dependencies No The paper does not provide specific version numbers for any software dependencies or libraries used in the implementation or experiments.
Experiment Setup No The paper states that "hyperparameters of all optimizers are tuned to achieve the best performance," but it does not provide specific values for these hyperparameters (e.g., learning rate, batch size, number of epochs) or detailed training configurations.