Adaptive Estimation of Graphical Models under Total Positivity

Authors: Jiaxi Ying, José Vinı́cius De Miranda Cardoso, Daniel P. Palomar

ICML 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on synthetic and financial time-series data demonstrate that our method outperforms state-of-the-art methods, concurrently achieving lower precision matrix estimation error and higher graph edge selection accuracy.
Researcher Affiliation Academia 1The Hong Kong University of Science and Technology, Clear Water Bay, Hong Kong 2HKUST Shenzhen-Hong Kong Collaborative Innovation Research Institute, Shenzhen, China.
Pseudocode Yes Algorithm 1 Solve Problem (5) and Algorithm 2 Compute PSA SC(Y )
Open Source Code Yes The code of our method is publicly available at https://github.com/jxying/ddmtp2.
Open Datasets Yes We conduct experiments on synthetic and financial time-series data.
Dataset Splits No No explicit percentages, absolute sample counts, or predefined splits for training, validation, and test sets are provided in the paper. The paper mentions synthetic data generation and financial time-series data collection with sample sizes, but not specific partitioning into train/validation/test sets.
Hardware Specification No No specific hardware details (e.g., GPU/CPU models, memory, or detailed computer specifications) used for running experiments are provided.
Software Dependencies No The paper mentions that code is publicly available, but it does not explicitly list specific software dependencies with version numbers (e.g., Python, PyTorch, CUDA versions) within the text.
Experiment Setup No The paper describes the methods and general data generation process, but it does not provide specific numerical values for hyperparameters (e.g., learning rate, batch size, number of epochs, optimizer settings) or detailed system-level training configurations within the main text.