SESSION AREA 01
LEARNING THEORY & OPTIMISATION
Generalisation, statistical learning, optimisation methods, uncertainty, and the mathematical foundations of reliable learning.
01
ML FORWARD SUMMIT
Programme
Explore machine learning research, systems and applications across 16 topic areas.
February 2-3, 2027 / Berlin, Germany
Explore four programme themes
Connect the 16 topic areas through prediction, model lifecycle engineering, efficient deployment and applied case studies.
- Applied ML for Prediction and Decisions
- Compare forecasting, ranking and anomaly detection through baselines, validation and the cost of prediction errors.
- MLOps and Model Lifecycle Engineering
- Examine data and model versioning, release controls, monitoring, rollback and defensible retraining decisions.
- Efficient Models and Edge Intelligence
- Compare compression and deployment choices against accuracy, memory, latency and maintenance constraints.
- Production ML Case Studies
- Discuss measured results, deployment failures and lessons from applied systems. Case studies connect across all 16 topic areas.
SESSION AREA 02
DEEP LEARNING & REPRESENTATION LEARNING
Neural architectures, self-supervised learning, transfer learning, feature representations, and training methods.
02
SESSION AREA 03
NLP & LARGE LANGUAGE MODELS
Language understanding, speech, retrieval-augmented generation, adaptation, reasoning, and language-model evaluation.
03
SESSION AREA 04
COMPUTER VISION & PERCEPTION
Image and video understanding, detection, segmentation, 3D vision, and visual learning across real-world environments.
04
SESSION AREA 05
GENERATIVE & MULTIMODAL LEARNING
Generative models, diffusion methods, cross-modal representations, and learning across text, images, audio, and video.
05
SESSION AREA 06
REINFORCEMENT LEARNING & DECISION-MAKING
Sequential decisions, offline and model-based reinforcement learning, planning, multi-agent learning, and policy evaluation.
06
SESSION AREA 07
GRAPH LEARNING & RECOMMENDER SYSTEMS
Graph neural networks, relational data, knowledge graphs, ranking, recommendations, and personalisation.
07
SESSION AREA 08
TIME SERIES & FORECASTING
Temporal validation, leakage prevention, forecasting with limited history, anomaly detection, and learning from changing demand.
08
SESSION AREA 09
CAUSAL & PROBABILISTIC MACHINE LEARNING
Causal discovery, causal inference, Bayesian methods, uncertainty estimation, and robust decision support.
09
SESSION AREA 10
DATA-CENTRIC ML & SYNTHETIC DATA
Data quality, labelling, synthetic data, feature freshness, training-serving consistency, and reproducible evaluation datasets.
10
SESSION AREA 11
MLOPS & PRODUCTION ML SYSTEMS
Reproducible pipelines, model registries, release gates, rollback, drift signals, retraining decisions, and maintenance ownership.
11
SESSION AREA 12
EFFICIENT ML & EDGE AI
Compression, quantisation, distillation, hardware-aware learning, and measuring accuracy, memory and end-to-end latency together.
12
SESSION AREA 13
FEDERATED & PRIVACY-PRESERVING LEARNING
Distributed learning, differential privacy, secure aggregation, and collaboration across sensitive datasets.
13
SESSION AREA 14
EVALUATION, SAFETY & RESPONSIBLE ML
Robustness, fairness, explainability, security, alignment, and evidence-based evaluation of machine learning systems.
14
SESSION AREA 15
ML FOR SCIENCE, HEALTH & CLIMATE
Scientific discovery, healthcare, biology, climate modelling, sustainability, and rigorous domain-specific validation.
15
SESSION AREA 16
ROBOTICS & EMBODIED LEARNING
Learning for perception, control, manipulation, autonomous systems, simulation, and interaction with the physical world.
16
CONTRIBUTE TO THE CONVERSATION.Explore research themes and contribution guidance.
CALL FOR PAPERSGO