Structural Hawkes Processes for Learning Causal Structure from Discrete-Time Event Sequences

Authors: Jie Qiao, Ruichu Cai, Siyu Wu, Yu Xiang, Keli Zhang, Zhifeng Hao

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

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on synthetic and real-world data verify the effectiveness of the proposed method.
Researcher Affiliation Collaboration Jie Qiao1 , Ruichu Cai1,2 , Siyu Wu1 , Yu Xiang1 , Keli Zhang3 and Zhifeng Hao4 1School of Computer Science, Guangdong University of Technology, Guangzhou 510006, China 2Peng Cheng Laboratory, Shenzhen 518066, China 3Huawei Noah s Ark Lab, Huawei, Shenzhen 518116, China 4College of Science, Shantou University, Shantou 515063, China
Pseudocode Yes Algorithm 1 Learning causal structure using SHP
Open Source Code No The paper does not contain any explicit statement or link indicating that the source code for the proposed methodology is publicly available.
Open Datasets Yes We also test the proposed SHP on a very challenging real-world dataset1 from real telecommunication networks. The dataset records eight months of alarms that occurred in a real metropolitan cellular network. 1https://competition.huaweicloud.com/informations/mobile/1000041487/dataset
Dataset Splits No The paper conducts experiments on synthetic and real-world data but does not explicitly provide specific details about training, validation, or test dataset splits (e.g., percentages or sample counts) that would be needed for reproduction.
Hardware Specification No The paper does not provide specific details about the hardware (e.g., GPU models, CPU types, memory) used to run the experiments.
Software Dependencies No The paper mentions methods and algorithms but does not specify any software dependencies with version numbers (e.g., Python, PyTorch, or specific libraries with their versions).
Experiment Setup No The paper describes the default settings for synthetic data generation (sample size, time interval, number of event types, ranges of alpha and mu), but it does not provide specific hyperparameters or system-level training settings for the model itself (e.g., learning rate, batch size, optimizer configuration).