Causal Reasoning in the Presence of Latent Confounders via Neural ADMG Learning
Authors: Matthew Ashman, Chao Ma, Agrin Hilmkil, Joel Jennings, Cheng Zhang
ICLR 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We further validate our approach via experiments on both synthetic and real-world datasets, and demonstrate the competitive performance against relevant baselines. and Empirical evaluation on synthetic and real-world datasets (Section 6). We evaluate N-ADMG on a variety of synthetic and real-world datasets, comparing performance with a number of existing causal discovery and inference algorithms. |
| Researcher Affiliation | Collaboration | Matthew Ashman2 , Chao Ma1 , Agrin Hilmkil1, Joel Jennings1, Cheng Zhang1 1Microsoft Research Cambridge 2University of Cambridge mca39@cam.ac.uk , {chaoma,agrinhilmkil,joeljennings,chezha}@microsoft.com |
| Pseudocode | No | The paper does not contain any clearly labeled 'Algorithm' or 'Pseudocode' blocks. |
| Open Source Code | Yes | Our implementation will be available at https://github.com/microsoft/causica. and First, we open source our package in our github page github.com/microsoft/causica/tree/v0.0.0. |
| Open Datasets | Yes | The outcomes of treatments are simulated artificially as in Hill (2011); hence the outcomes of both treatments (home visits or not) on each subject are known. ... We use 10 replicates of different simulations based on setting B (log-linear response surfaces) of Hill (2011), which can downloaded from https://github.com/AMLab-Amsterdam/CEVAE. and All datasets used in this paper are either publicly available data, or synthetic data whose generation process is described in detail. |
| Dataset Splits | No | We use a 70%/30% train-test split ratio. Before training our models, all continuous covariates are normalised. The paper only mentions a train-test split and does not explicitly describe a separate validation split or its proportion. |
| Hardware Specification | No | The paper does not provide any specific details about the hardware (e.g., GPU/CPU models, memory, cloud instance types) used to run the experiments. |
| Software Dependencies | No | The paper mentions software components and techniques like 'Adam', 'MLPs', 'Gumbel softmax method', and 'variational inference' but does not specify any programming languages, libraries, or solvers with version numbers. |
| Experiment Setup | Yes | We use a learning rate of 10 3 for the model parameters and 5 10 3 for the variational parameters. We optimise the objective for a maxmimum of 5000 steps or until convergence (we stop early if the loss does not improve for 1500 optimisation steps...) and The functions ξ1, ξ2 and ℓused in the likelihood and the inference network used to parameterise qϕ(u|x) are all two hidden layer MLPs with 80 hidden units per hidden layer. |