Causal structure-based root cause analysis of outliers

Authors: Kailash Budhathoki, Lenon Minorics, Patrick Bloebaum, Dominik Janzing

ICML 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We study the empirical performance of the method through simulations and present a real-world case study identifying root causes of extreme river flows.
Researcher Affiliation Industry 1Amazon Research Tübingen. Correspondence to: Kailash Budhathoki <kaibud@amazon.com>.
Pseudocode No The paper does not contain any clearly labeled pseudocode or algorithm blocks.
Open Source Code Yes The implementation is available from the gcm module (Bl obaum et al., 2022) in Do Why.
Open Datasets Yes Data Source: https://tinyurl.com/ukriverdata
Dataset Splits No The paper only mentions 'training samples' and 'test samples' but does not specify a separate validation split or explicit validation methodology.
Hardware Specification No The paper does not provide specific details about the hardware used for running the experiments (e.g., CPU, GPU models, memory).
Software Dependencies No The paper mentions 'gcm module... in Do Why' but does not provide specific version numbers for software dependencies (e.g., Python, PyTorch, or the mentioned modules).
Experiment Setup Yes To each node Xj, we assign a random linear structural equation of the form Xj := P i βij PAij + Nj, where PAij is the i-th component of Xj s parents PAj, βij Uniform(0, 5) and Nj Gaussian(0, 1). ... With a threshold of z = 3, we identify four outliers.