Data Science
Data Science Group
Statistical Machine Learning and Artificial Intelligence
Our Data Science Group pursues foundational research across statistical machine learning and artificial intelligence, including reinforcement learning, computer vision, and natural language processing. This research aims to progress core capabilities in perception, reasoning, prediction and decision-making. When integrated thoughtfully into multidisciplinary systems, our fundamentally-driven AI can enable autonomous agents, enhance imaging-based diagnoses, optimize industrial processes, and unlock insights from unstructured data at new scales. We collaborate with domain experts in engineering, healthcare, manufacturing, and more to responsibly apply our machine learning innovations to address real-world challenges. Our goal is advancing the scientific frontier to expand what machines can perceive, understand and act upon.
Key contact
Giovanni Montana - Professor
Business resource
Key links
Focus Areas
AI for Autonomous Agents
Our research focuses on using reinforcement learning (RL) to develop autonomous artificial intelligence agents capable of complex decision-making. We explore deep neural network architectures to enable single agents to perceive their environment and learn optimal policies for tasks from raw sensory inputs. Additionally, we study multi-agent RL systems and decentralized training methods like networked agents and emergent communication protocols, which allow potentially large numbers of AI agents to coordinate behaviours and achieve shared goals. This work on scalable, decentralized autonomous agents has applications from distributed logistics to manufacturing robotics and optimization of digital twins for smart cities.
AI in Digital Healthcare
Our research applies artificial intelligence, especially deep learning, to advance digital healthcare services. A major focus is developing deep neural networks for medical image analysis that can assist diagnosis through automated detection and classification of anatomical structures and abnormalities in radiology scans. Additionally, we utilize large language models to extract insights from radiological reports and other clinical notes. Beyond imaging, we also study integrations of AI into real-world NHS workflows, working closely with hospitals to responsibly deploy systems that improve diagnostic efficiency, personalize treatment, and enable more accurate prognoses and disease monitoring.
AI for Image and Video Understanding
Our research focuses on advancing artificial intelligence capabilities for understanding visual content, including both static images and video. We develop deep neural networks to perform computer vision tasks such as object detection, semantic segmentation, and anomaly detection. A major emphasis is placed on video analytics, where we apply techniques like recurrent models and 3D convolutions to enable detailed spatio-temporal understanding. This research aims to progress machine perception and interpretation of video content. Specific opportunities include the development of autonomous agents that can sense and interpret the visual world, high-level indexing and retrieval of video databases, video captioning, and more. The goal is to unlock the rich information stored in visual media.
Ready to work with WMG?
Register your interest in our Data Science research to start the conversation with us.
Your content here
Your content here