FMLTS 2026

Future of Machine Learning for Time Series Workshop @ ACML 2026
1 December 2026, Melbourne

About

Machine learning for time series is in a moment of transition. Traditional specialised methods are now joined by deep learning and foundation models. Yet most work is still done using very small datasets, and a significant amount of work suffers from weaknesses in benchmarking and evaluation practices.

This workshop will bring together researchers, students, and practitioners from across disciplines to explore future directions and engage with the most pressing challenges in machine learning research for time series.

Topics

We invite submissions on the following topics in relation to machine learning for time series:

• time series classification
• forecasting
• anomaly detection
• extrinsic regression
• datasets, benchmarking, and generalisation
• bias and variance
• inductive bias and ‘no free lunch’
• deep learning and foundation models
• embeddings and representations
• scalability
• segmentation

Submissions

Submissions should follow the ACML LaTeX template and style file (zip).

Submissions should be from 4 to 6 pages in length (excluding references and appendices).

Submissions must be anonymised.

Submissions via OpenReview.†

We aim to have each submission reviewed by three reviewers. Accepted papers will be delivered either as oral presentations or posters.

^{†OpenReview moderation of new profiles without an institutional email address can take up to two weeks.}

Important Dates

18 October 2026	Submission Deadline
30 October 2026	Outcome Notification
13 November 2026	Camera Ready Deadline
1 December 2026	Workshop

Program

TBC

Speakers

Professor James Bailey
Head of Department, Data Science and Artificial Intelligence
Monash University

TBC

Organisers

Angus Dempster Monash University

Yunrui Zhang Monash University

Chang Wei Tan Oracle

Contact

Angus Dempster
angus.dempster@monash.edu