Dr Deyu Ming, Lecturer in Mathematics and Data Analytics at the UCL School of Management, has won the 2025 UCL Open Science & Scholarship Award for his work as lead developer of dgpsi, an advanced software that has helped to support AI projects at the forefront of sustainable solutions.
Dr Ming received the award in the Open-Source Software / Analytical Tools category, which recognises his outstanding contribution to the promotion of open science through the development of the open-source software package, dgpsi.
dgpsi, which stands for “deep Gaussian processes using stochastic imputation”, is designed to build high-performance, trustworthy and reusable machine-learning models for fast and scalable predictions.
The software is made available in both R and Python programming languages and freely accessible on public repositories such as GitHub, PyPI, Conda-forge and CRAN. Since its launch in 2021 dgpsi has surpassed 100,000 downloads worldwide across multiple operating systems, with 19 public releases.
When it comes to real-world usage, the software has helped underpin ADD-TREES, an AI research project that enables clients such as Defra (Department for Environment, Food & Rural Affairs) to design tree-planting strategies for carbon sequestration, while meeting biodiversity targets within land and budget constraints.
Powered by hundreds of models built with dgpsi, the ADD-TREES App can generate more than 300 million predictions in under one minute, delivering real-time exploration of land-use scenarios.
Speaking about the award, Dr Ming said:
“I’m deeply honoured to receive this award and recognition. I feel very fortunate to be working in an era shaped by strong momentum in open science and a growing global commitment to open-source software, an environment in which dgpsi has received invaluable input from its users and the wider open-source community.
“Thanks to their contributions, the package has evolved remarkably over the past years, achieving greater robustness, more than a 60-fold speed-up and tangible real-world impact in helping to address Net Zero challenges.
“In the years ahead, I hope it will continue to become faster, lighter, smarter and more open, with the support of the open-source community, so it remains accessible, transparent, reproducible and useful to an ever-wider range of users and impactful applications.”
Learn more about dgpsi and the ADD-TREES project:
