Causal Discovery from Discrete Data using Hidden Compact Representation
Authors: Ruichu Cai, Jie Qiao, Kun Zhang, Zhenjie Zhang, Zhifeng Hao
NeurIPS 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Empirical studies verify the effectiveness of the proposed approach on both synthetic and real-world data. 4 Experiments To investigate the effectiveness of the proposed method based on the hidden compact representation model, we compare it with baseline algorithms on both synthetic data and the real world data. |
| Researcher Affiliation | Collaboration | Ruichu Cai 1, Jie Qiao1, Kun Zhang2, Zhenjie Zhang3, Zhifeng Hao1, 4 1 School of Computer Science, Guangdong University of Technology, China 2 Department of philosophy, Carnegie Mellon University 3 Singapore R&D, Yitu Technology Ltd. 4 School of Mathematics and Big Data, Foshan University, China |
| Pseudocode | Yes | Algorithm 1 Optimization of max L = supf maxθ L |
| Open Source Code | Yes | The implementation of HCR can be found on CRAN 1. 1https://cran.r-project.org/package=HCR |
| Open Datasets | Yes | On real-world data, we run the algorithm on Pittsburgh Bridges dataset and Abalone dataset. Both of them are wildly used in previous research and can be downloaded from UCI Machine Learning Repository [Lichman, 2013]. |
| Dataset Splits | No | The paper mentions sample sizes and the generation process for synthetic data but does not provide specific train/validation/test splits, percentages, or absolute sample counts for each split in a reproducible manner. |
| Hardware Specification | No | The paper does not specify any hardware details (e.g., CPU, GPU models, memory) used for running the experiments. |
| Software Dependencies | No | The paper mentions that the HCR implementation is on CRAN (implying R), but it does not provide specific version numbers for R or any dependent libraries or software. |
| Experiment Setup | Yes | In this set of experiments, the samples are generated according to the following two-stage procedure. Firstly, generate X from a multinomial distribution and its cardinality is randomly chosen from {3, 4, ..., 15}. Secondly, map each X to a value that uniformly samples from the interval {1, 2, ..., X }. Finally, randomly generate a conditional probability distribution P(Y Y ) and sample Y according to Y and P(Y Y ), and Y is generated from the interval { Y , ..., 15}. |