Disentangled Representation for Causal Mediation Analysis

Authors: Ziqi Xu, Debo Cheng, Jiuyong Li, Jixue Liu, Lin Liu, Ke Wang

AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results show that the proposed method outperforms existing methods and has strong generalisation ability. We further apply the method to a real-world dataset to show its potential application. We evaluate the effectiveness of the DMAVAE method on synthetic datasets. Experiments show that DMAVAE outperforms existing CMA methods and has a strong generalisation ability.
Researcher Affiliation Academia Ziqi Xu1, Debo Cheng1,2*, Jiuyong Li1*, Jixue Liu1, Lin Liu1, Ke Wang3 1 University of South Australia 2 Guangxi Normal University 3 Simon Fraser University
Pseudocode No The paper includes 'Figure 3: The architecture of Disentangled Mediation Analysis Variational Auto Encoder (DMAVAE)' and mathematical equations for the model. However, it does not contain any formal pseudocode or algorithm blocks.
Open Source Code Yes The code is available at https://github.com/IRON13/DMAVAE.
Open Datasets Yes We apply DMAVAE on the real-world dataset Adult. The dataset is retrieved from the UCI repository (Dua and Graff 2017) and it contains 14 attributes including personal, educational and economic information for 48842 individuals.
Dataset Splits No The paper states that synthetic datasets are generated with various sample sizes (2k-10k) and 30 repetitions, and a real-world dataset (Adult) is used. However, it does not explicitly provide details about the train/validation/test splits for these datasets (e.g., specific percentages or sample counts for each split).
Hardware Specification No The paper does not provide any specific details about the hardware (e.g., CPU, GPU models, memory, or processing units) used to run the experiments.
Software Dependencies No The paper mentions using Py Torch, Pyro, and R packages like 'mediation', 'Medflex', and 'causal-weight'. However, it does not provide specific version numbers for these software components, which are necessary for reproducible dependency descriptions.
Experiment Setup No The paper describes aspects of the data generation setup, such as sample sizes for synthetic datasets (2k-10k) and the number of repetitions (30). However, it does not provide specific details regarding the hyperparameters or training configurations for the DMAVAE model itself (e.g., learning rate, batch size, number of epochs, optimizer settings).