Causal Attribution Analysis for Continuous Outcomes
Authors: Shanshan Luo, Yu Yixuan, Chunchen Liu, Feng Xie, Zhi Geng
ICML 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | An artificial hypertension example and a real developmental toxicity dataset are employed to illustrate our method. ... To assess the stability of the proposed estimation procedure in Section 4 of the Supplementary Material, we conduct simulation studies by generating data according to the causal network depicted in Figure 1. The estimated results of Table 2 under different sample sizes were provided. |
| Researcher Affiliation | Collaboration | 1School of Mathematics and Statistics, Beijing Technology and Business University, Fangshan District, Beijing, China 2Ling Yang Co.Ltd, Alibaba Group, Hangzhou, China. |
| Pseudocode | No | The paper describes a "two-step estimation procedure" and identifies estimands using mathematical expressions and lemmas, but it does not present any structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not contain an explicit statement about releasing code, nor does it provide a link to a code repository. The supplementary material description mentions "simulation details of the proposed procedure" but does not explicitly state that code is provided. |
| Open Datasets | Yes | In Section S10 of the Supplementary Material, we also apply the proposed method to a real dataset (NTP, 2023). ... from the developmental toxicology experiments conducted by the National Toxicology Program (NTP) (NTP, 2023). ... https://cebs.niehs.nih.gov/cebs/publication/TR-602 |
| Dataset Splits | No | The simulation studies mention using "2000000 samples" for generating true values and then running simulations with "n = 1000, 2000, and 10000" samples over "500 repetitions". For the real dataset, it states "a total of 120 mice were randomly exposed to six different dose levels". This describes sample sizes and data generation, but not specific train/test/validation splits for machine learning experiments. |
| Hardware Specification | No | The paper does not provide any specific details about the hardware used to run the experiments, such as CPU or GPU models, memory, or specific cloud instance types. |
| Software Dependencies | No | The paper mentions using "the R package bnlearn to construct a Bayesian network" in the real data analysis section, but it does not specify any version numbers for R or the bnlearn package. This is insufficient for reproducibility. |
| Experiment Setup | No | The paper describes the theoretical framework, identifiability, and estimation methods for causal estimands. While it presents simulation results and real data analysis, it does not explicitly provide concrete experimental setup details such as hyperparameters, learning rates, batch sizes, or specific training configurations for any models used in estimation. |