Causal Domain Adaptation

Causal domain adaptation focuses on improving the performance of machine learning models when transferring knowledge from a data-rich source domain to a data-scarce target domain by explicitly modeling causal relationships between variables. Current research emphasizes developing algorithms that identify and leverage invariant causal mechanisms across domains, often employing techniques like counterfactual inference and conditional independence tests to discover and exploit these relationships. This approach is proving valuable in diverse applications, including improving the reasoning capabilities of large language models, enhancing industrial time-series forecasting, and enabling more robust analysis of financial data by identifying causal factors driving corporate performance changes.

Papers

September 4, 2024

Deconfounded Causality-aware Parameter-Efficient Fine-Tuning for Problem-Solving Improvement of LLMs
Ruoyu Wang, Xiaoxuan Li, Lina Yao
Large Language Model Reasoning Capability Large Relevance Improvement Causal Domain Adaptation

July 19, 2024

Domain Adaptation for Industrial Time-series Forecasting via Counterfactual Inference
Chao Min, Guoquan Wen, Jiangru Yuan, Jun Yi, Xing Guo
Domain Adaptation Domain Invariant Representation Counterfactual Inference Industrial Time Series Causal Domain Adaptation

October 24, 2023

CR-COPEC: Causal Rationale of Corporate Performance Changes to Learn from Financial Reports
Ye Eun Chun, Sunjae Kwon, Kyunghwan Sohn, Nakwon Sung, Junyoup Lee, Byungki Seo, Kevin Compher, Seung-won Hwang, Jaesik Choi
MAESTRO Dataset Causal Direction Causal Analysis Causal Reasoning Financial Report Causal Domain Adaptation

February 27, 2022

Causal Domain Adaptation with Copula Entropy based Conditional Independence Test
Jian Ma
Domain Adaptation Causal Discovery Causal Graph Conditional Independence Dimensional Copula Causal Domain Adaptation

Causal Domain Adaptation

Papers

Deconfounded Causality-aware Parameter-Efficient Fine-Tuning for Problem-Solving Improvement of LLMs

Domain Adaptation for Industrial Time-series Forecasting via Counterfactual Inference

CR-COPEC: Causal Rationale of Corporate Performance Changes to Learn from Financial Reports

Causal Domain Adaptation with Copula Entropy based Conditional Independence Test