Causal Inference with Conditional Front-Door Adjustment and Identifiable Variational Autoencoder
Authors: Ziqi Xu, Debo Cheng, Jiuyong Li, Jixue Liu, Lin Liu, Kui Yu
ICLR 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on synthetic datasets validate the effectiveness of CFDi VAE and its superiority over existing methods. We further apply CFDi VAE to a real-world dataset to demonstrate its potential application. |
| Researcher Affiliation | Academia | Ziqi Xu 1,2, Debo Cheng 1, , Jiuyong Li 1, , Jixue Liu 1, Lin Liu 1 & Kui Yu 3 University of South Australia 1 Data61, CSIRO 2 Hefei University of Technology 3 |
| Pseudocode | Yes | Algorithm 1 CFD Adjustment for ATE Estimation with Discrete Data |
| Open Source Code | Yes | The source code is available in the Supplementary Material. |
| Open Datasets | Yes | In this section, we apply CFDi VAE to detect discrimination on the real-world dataset, Adult. The dataset is retrieved from the UCI repository (Dua & Graff, 2017) and it contains 11 attributes about personal, educational and economic information for 48842 individuals. |
| Dataset Splits | No | The paper describes generating datasets of various sizes for evaluation and running methods multiple times, but it does not specify how these datasets are partitioned into explicit training, validation, and test sets for model development and evaluation. |
| Hardware Specification | No | The paper does not provide specific hardware details such as GPU or CPU models, memory, or cloud instance types used for running the experiments. |
| Software Dependencies | No | The paper states "CFDi VAE is implemented in Python (Van Rossum & Drake Jr, 1995) libraries Py Torch (Paszke et al., 2019) and Pyro (Bingham et al., 2019). The code for data generation is written in R (R Core Team, 2021)." However, it does not provide specific version numbers for PyTorch, Pyro, or other libraries used beyond the base languages. |
| Experiment Setup | Yes | The paper provides a table (Table 6) detailing parameter settings for CFDi VAE, including Reps 30, Epoch 30, Batch Size 256, Num Layers 3, lr 1e-3, lrd 0.01, and wd 1e-4. |