Based Causal Discovery
Based causal discovery aims to infer causal relationships from observational data, focusing on developing robust algorithms that accurately reconstruct causal graphs despite complexities like hidden confounders and noisy data. Current research emphasizes improving score-based methods, often employing kernel-based approaches or neural networks, and incorporating prior knowledge to constrain the search space and enhance efficiency. These advancements are crucial for reliable causal inference across diverse scientific domains and applications, enabling better understanding of complex systems and more effective decision-making.
Papers
July 18, 2024
July 14, 2024
July 8, 2024
October 27, 2023
May 26, 2023
April 11, 2023
March 6, 2023