Bio
Joseph completed an undergraduate degree in computer science at the University of Southampton, and then specialised with a master’s degree focused on machine learning (also at the University of Southampton). Joseph pursed his interest in machine learning and is now researching explainable artificial intelligence – the process of unpacking the “black box” of machine learning and understanding why models exhibit certain behaviours.
Joseph is motivated by the possibility for collaboration within the Turing, with a strong interest in helping others apply and understand machine learning in their domains. He believes a drive towards openness, transparency and explainability in artificial intelligence will allow the recent advances in machine learning to spread and contribute to other areas such as medicine and urban analytics.
Joseph also has experience working in the start-up industry. Using his specialist machine learning knowledge, he contributed to a start-up within the University of Southampton during his master’s degree. He is excited to see how the field of artificial intelligence will progress in the future and is looking forward to working alongside other researchers from a diverse range of fields.
Research interests
Joseph’s doctoral research has investigated how the decision-making process of machine learning models can be better understood. they have specifically focused on multiple instance learning, where models learn from inexact labels, and developed a range of interpretability techniques for use in computer vision, reinforcement learning, and time series analysis.
Selected publications and papers
Joseph Early, Christine Evers, Sarvapali Ramchurn. "Model Agnostic Interpretability for Multiple Instance Learning". ICLR 2022
Joseph Early, Tom Bewley, Christine Evers, Sarvapali Ramchurn. "Non-Markovian Reward Modelling from Trajectory Labels via Interpretable Multiple Instance Learning". NeurIPS 2022
Joseph Early, Ying-Jung Deweese, Christine Evers, Sarvapali Ramchurn. "Scene-to-Patch Earth Observation: Multiple Instance Learning for Land Cover Classification". Tackling Climate Change with Machine Learning: Workshop at NeurIPS 2022
Gregory Everett, Ryan Beal, Tim Matthews, Joseph Early, Timothy Norman, Sarvapali Ramchurn. "Inferring Player Location in Sports Matches: Multi-Agent Spatial Imputation from Limited Observations". AAMAS 2023
Keri Grieman, Joseph Early. "A Risk-based Approach to AI Regulation: System Categorisation and Explainable AI Practices". SCRIPTed Journal 2023.
Saqib Rahman, Joseph Early, Matt De Vries, Megan Lloyd, Ben Grace, Gopal Ramchurn, Timothy Underwood. "Predicting response to neoadjuvant therapy using image capture from diagnostic biopsies of oesophageal adenocarcinoma". European Journal of Surgical Oncology 2022