{"database":"biostudies-literature","file_versions":[],"scores":null,"additional":{"omics_type":["Unknown"],"volume":["20(10)"],"submitter":["Wheeler J"],"pubmed_abstract":["Autoregressive moving average (ARMA) models are widely used for analyzing time series data. However, standard likelihood-based inference methodology for ARMA models has avoidable limitations. We show that currently accepted standards for ARMA likelihood maximization frequently lead to sub-optimal parameter estimates. Existing algorithms have theoretical support, but can result in parameter estimates that correspond to a local optimum. While this possibility has been previously identified, it remains unknown to most users, and no routinely applicable algorithm has been developed to resolve the issue. We introduce a novel random initialization algorithm, designed to take advantage of the structure of the ARMA likelihood function, which overcomes these optimization problems. Additionally, we "],"journal":["PloS one"],"pagination":["e0333993"],"full_dataset_link":["https://www.ebi.ac.uk/biostudies/studies/S-EPMC12551883"],"repository":["biostudies-literature"],"pubmed_title":["Revisiting inference for ARMA models: Improved fits and superior confidence intervals."],"pmcid":["PMC12551883"],"pubmed_authors":["Ionides EL","Wheeler J"],"additional_accession":[]},"is_claimable":false,"name":"Revisiting inference for ARMA models: Improved fits and superior confidence intervals.","description":"Autoregressive moving average (ARMA) models are widely used for analyzing time series data. However, standard likelihood-based inference methodology for ARMA models has avoidable limitations. We show that currently accepted standards for ARMA likelihood maximization frequently lead to sub-optimal parameter estimates. Existing algorithms have theoretical support, but can result in parameter estimates that correspond to a local optimum. While this possibility has been previously identified, it remains unknown to most users, and no routinely applicable algorithm has been developed to resolve the issue. We introduce a novel random initialization algorithm, designed to take advantage of the structure of the ARMA likelihood function, which overcomes these optimization problems. Additionally, we ","dates":{"release":"2025-01-01T00:00:00Z","publication":"2025","modification":"2026-06-05T04:31:31.938Z","creation":"2026-05-13T14:24:54.356Z"},"accession":"S-EPMC12551883","cross_references":{"pubmed":["41134851"],"doi":["10.1371/journal.pone.0333993"]}}