CausalNET: Unveiling Causal Structures on Event Sequences by Topology-Informed Causal Attention
Authors: Hua Zhu, Hong Huang, Kehan Yin, Zejun Fan, Hai Jin, Bang Liu
IJCAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Comprehensive evaluation on a spectrum of realworld and synthetic datasets underscores the superior performance and scalability of Causal NET, marking a promising step forward in the realm of causal discovery. Code and Appendix are available at https://github.com/CGCL-codes/Causal NET. |
| Researcher Affiliation | Academia | Hua Zhu1 , Hong Huang1 , Kehan Yin1 , Zejun Fan1 , Hai Jin1 and Bang Liu2 1National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab, Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology, Wuhan, China 2DIRO, Universit e de Montr eal & Mila & Canada CIFAR AI Chair, Canada {huazhu, honghuang, kehanyin, zejunfan, hjin}@hust.edu.cn, bang.liu@umontreal.ca |
| Pseudocode | No | The paper describes the model and training procedure using text and mathematical equations but does not provide structured pseudocode or an algorithm block. |
| Open Source Code | Yes | Code and Appendix are available at https://github.com/CGCL-codes/Causal NET. |
| Open Datasets | Yes | We adopt two challenging real-world datasets from telecommunication networks 2. The first is 24V 439N Microwave (Micro-24)... The second is 25V 474N Microwave (Micro-25)... ... To explore the applicability of our model in other domains ... we further include the IPTV dataset [Luo et al., 2015]... In addition, we also generate a range of synthetic datasets via gcatle s API [Zhang et al., 2021]... 2https://competition.huaweicloud.com/information/1000041487/ dataset |
| Dataset Splits | No | The paper discusses training and testing but does not explicitly provide details on how the datasets were split into training, validation, and test sets (e.g., percentages or specific methods). |
| Hardware Specification | Yes | TNPAR encounters out-of-memory (OOM) issues on datasets with 100 and 150 event types on Tesla V100 GPUs (32 GB). |
| Software Dependencies | No | The paper does not explicitly provide a list of software dependencies with specific version numbers. |
| Experiment Setup | No | The paper mentions `hyper-parameter ξ` (max time lag) and `hyper-parameter k` (max hop) and discusses their influence, as well as `temperature parameter τ` for Gumbel Softmax. However, it does not provide a comprehensive list of other common hyperparameters (e.g., learning rate, batch size, number of epochs, optimizer settings) or other specific system-level training configurations. |