21 January 2026, Singapore Expo,
Room Garnet 214 (second floor)
09:00 - 09:15
09:15 - 10:00
10:00 - 10:30
Coffee Break 10:30 - 11:00
11:00 - 11:15
11:15 - 11:45
Authors of accepted papers will give a short spotlight presentation introducing their work.
Archival Track (2 minutes and 30 seconds):
Paper 2: Nitin Vetcha: "Human-Pedagogy Inspired LLM Fine-Tuning Paradigm for Lifelong Leaning and Continual Adaptation"
Paper 3: Nitin Vetcha, Dianbo Liu: "SOLAR : A Self-Optimizing Open-Ended Autonomous Agent for Lifelong Learning and Continual Adaptation"
Paper 13: Lorenzo Iovine, Giacomo Ziffer, Emanuele Della Valle: "Tracking Adaptation Time: Metrics for Temporal Distribution Shift"
Paper 17: Tahir Qasim Syed, Behraj Khan: "AdaShiftBoost: Distribution-Aware Gradient Boosting under Covariate Shift"
Paper 18: Marcello Matteo Declich: "Towards Streaming Continual Learning for Earth Observation Multimodal Foundation Models"
Paper 20: Afonso Lourenço, Joao Gama, Eric P. Xing, Goreti Marreiros: "Bridging Streaming Continual Learning via In-Context Large Tabular Models"
Paper 22: Yuantao Fan, Slawomir Nowaczyk: "Online Learning for Energy Consumption Forecasting in Heavy-Duty Electric Vehicles"
Paper 23: Aurora Esteban, Sepideh Pashami, Slawomir Nowaczyk: "Online Learning Supported by Foundation Models for Anomaly Detection in Industrial Settings"
Non-archival Track (1 minute and 30 seconds):
Paper 1: Edoardo Urettini, Daniele Atzeni, Ioanna-Yvonni Tsaknaki, Antonio Carta: "Online Continual Learning for Time Series: a Natural Score-driven Approach"
Paper 15: Keonvin Park: "Bridging Streaming and Temporal Adaptation: Toward Continual Time Series Learning"
11:45 - 12:30
The session will take place in the designated poster area on the second floor near the lecture rooms.
Lunch Break: 12.30-14.00
14:00 - 15:00
Abstract: Environmental data is highly dynamic, with shifting distributions and continuous change that challenge traditional machine learning methods. The TAIAO programme offers a rich testbed for studying these issues through large-scale datasets on biodiversity, climate, and land use. This talk will present insights from TAIAO on how Streaming and Continual Learning (SCL) can enable adaptive, reliable models in non-stationary settings. I will highlight advances in drift-aware algorithms, online algorithms, and scalable open-source tools, along with examples of how adaptive AI supports environmental monitoring and decision-making.
Bio: Albert Bifet is Director of the AI Institute at the University of Waikato, and Professor at Telecom Paris, IP Paris. His research focuses on machine learning for data streams, concept drift, and continual learning, with an emphasis on scalable algorithms and open-source scientific software. He is co-leader of MOA, a widely used frameworks for mining evolving data. He has published extensively on adaptive learning methods, drift detection, and streaming evaluation protocols, and has led multidisciplinary projects applying these techniques in environmental science and other dynamic domains. His work aims to bridge fundamental advances in continual learning with high-impact real-world applications.
15:00 - 15:30
Coffee Break 15:30 - 16:00
16:00 - 16:55
16:55 - 17.00