Causal Inference and Mechanism Clustering of A Mixture of Additive Noise Models

Authors: Shoubo Hu, Zhitang Chen, Vahid Partovi Nia, Laiwan CHAN, Yanhui Geng

NeurIPS 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on synthetic and real data demonstrate the effectiveness of our proposed approach. In this section, experimental results on both synthetic and real data are given to show the performance of ANM-MM on causal inference and mechanism clustering tasks.
Researcher Affiliation Collaboration The Chinese University of Hong Kong; Huawei Noah s Ark Lab; Huawei Montréal Research Center {sbhu, lwchan}@cse.cuhk.edu.hk {chenzhitang2, vahid.partovinia, geng.yanhui}@huawei.com
Pseudocode Yes Algorithm 1: Causal Inference; Algorithm 2: Mechanism clustering
Open Source Code Yes The Python code of ANM-MM is available online at https: //github.com/amber0309/ANM-MM.
Open Datasets Yes Causal inference on Tüebingen cause-effect pairs. We evaluate the causal inference performance of ANM-MM on real world benchmark cause-effect pairs3 [15]. (Footnote 3: https://webdav.tuebingen.mpg.de/cause-effect/.) Clustering on BAFU air data. We evaluate the clustering performance of ANM-MM on real air data obtained online5. (Footnote 5: https://www.bafu.admin.ch/bafu/en/home/topics/air.html)
Dataset Splits No The paper describes synthetic data generation and sampling for real data, but it does not specify explicit train/validation/test splits (e.g., 80/10/10 split or k-fold cross-validation) for reproduction.
Hardware Specification No The paper does not provide any specific hardware details such as GPU models, CPU types, or memory specifications used for running the experiments.
Software Dependencies No The paper mentions 'Python code' but does not specify version numbers for Python or any other software libraries or dependencies.
Experiment Setup Yes ANM-MM was applied using different parameter λ {0.001, 0.01, 0.1, 1, 10} and IGCI was applied using different reference measures and estimators. To find Θ, we resort to the gradient descant methods.