Improving generalisation via anchor multivariate analysis

arXiv:2403.01865v1 Announce Type: cross
Abstract: We introduce a causal regularisation extension to anchor regression (AR) for improved out-of-distribution (OOD) generalisation. We present anchor-compatible losses, aligning with the anchor framework to ensure robustness against distribution shifts. Various multivariate analysis (MVA) algorithms, such as (Orthonormalized) PLS, RRR, and MLR, fall within the anchor framework. We observe that simple regularisation enhances robustness in OOD settings. Estimators for selected algorithms are provided, showcasing consistency and efficacy in synthetic and real-world climate science problems. The …

abstract algorithms analysis anchor arxiv cs.lg distribution extension framework losses multivariate observe regression robustness stat.ap stat.me stat.ml type via