v148: Workshop on Preregistration in Machine Learning

· Source: Proceedings of Machine Learning Research · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems · Depth: Expert, short

Summary

Volume 148 compiles papers presented at the NeurIPS 2020 Workshop on Pre-registration in Machine Learning, held virtually on December 11, 2020, and edited by Luca Bertinetto, João F. Henriques, Samuel Albanie, Michela Paganini, and Gül Varol. The workshop featured diverse research across various machine learning paradigms and applications. Key contributions include advancements in computer vision tasks such as point cloud co-segmentation, object detection, video generation, and skeleton action recognition, often leveraging self-supervised learning and neural differential equations. Other significant areas explored are meta-learning, reinforcement learning (including hybrid quantum-classical approaches), domain adaptation, and unified lifelong learning frameworks. Furthermore, papers address critical challenges like adversarial robustness, algorithmic fairness, low-rank matrix completion, and the performance of federated learning algorithms.

Key takeaway

The NeurIPS 2020 Workshop on Pre-registration in Machine Learning compiles 24 papers exploring diverse advancements across core ML domains. Key contributions include novel techniques in computer vision (e.g., point cloud segmentation, video generation), reinforcement learning (e.g., hybrid quantum-classical RL), federated learning, and addressing model robustness and fairness. This collection offers critical insights for researchers and practitioners aiming to enhance model reliability, efficiency, and tackle complex challenges like domain shift and ethical considerations.

Topics

Best for: AI Scientist, Machine Learning Engineer

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Proceedings of Machine Learning Research.