Ryan Chan

Ryan Chan

I'm currently a researcher at The Alan Turing Institute focusing on end-to-end machine-learning weather forecasting.

I have a background in statistics and machine learning, and hold a PhD in Statistics from the University of Warwick, supervised by Professor Gareth Roberts and Professor Murray Pollock . In my PhD, I focused on developing Monte Carlo methods for combining distributed statistical analyses (Fusion). I completed my undergraduate studies at the University of Leeds, where I was awarded a first class honours MMath Mathematics degree and recieved the Royal Statistical Society Prize.

Research & Projects

I am currently interested in applying machine learning to environment and sustainability problems. I have recently started to work on end-to-end machine-learning weather forecasting and foundation models for Earth systems.

• FastNet: AI for Numerical Weather Prediction (AI4NWP)
  • FastNet is a Graph Neural Network for medium-range weather forecasting, developed in collaboration between the Met Office and The Alan Turing Institute.

Previous projects

• prompto: supporting reproducible experiments on LLMs at scale
  • prompto is an open source software package developed to support reproducible experiments on Large Language Models at scale. The tool interacts with self-hosted open source models as well as proprietary API-based services, making use of individual rate and token limits, to optimise for large-scale evaluations. [NAACL 2025 System Demonstrations Paper].
• Rough Paths
  • Developed software and applications of Rough Path theory and path signatures - a mathematical tool for describing complex streams of data in a structured and standardised way.
  • Worked on applying path signatures in natural language processing problems with the DataSig research team.
    • nlpsig: Python library for constructing paths of embeddings obtained from transformers.
    • sig-networks: Python/PyTorch library for training signature neural networks for longitudinal NLP tasks. [EACL 2024 System Demonstrations Paper].
    • SigMahaKNN: Python implementation of an algorithm for anomaly detection on mulitvariate streams using path signatures and the variance norm. [Paper].
• IceNet
  • IceNet is a probabilistic, deep learning sea ice forecasting system developed by an international team and led by British Antarctic Survey and The Alan Turing Institute.
  • Contributed to the development a machine learning pipeline for operational use of IceNet.
  • IceNet was nominated for the Artic Circle's Frederik Paulsen Arctic Academic Action Award.

Miscellaneous projects

• Maya: Multimodal Aya (Cohere For AI's Expedition Aya 2024 project)
• I was part of the Maya project that won the Most Promising Award at C4AI's Expedition Aya challenge.
• We pre-trained and fine-tuned a multilingual multimodal model based on Cohere's Aya model and the LLaVA architecture. [CVPR 2025 Workshop (Vision Language Models For All) Paper].
• Reginald: A friendly Turing Slack bot
  • A side project born out the Research Engineering Team's Hackweek 2023 at the Turing.
  • Developed a LLM chatbot for the institute which uses Retreival Augmented Generation and llama-index with access to internal and publicly available data for the institute.
  • It was deployed on the Turing's Slack workspace, as well as a command-line interface to interact with the model directly and locally on the terminal.
• AIrsenal & AIrgentina
  • AIrsenal is a Python package for using Machine learning to pick a Fantasy Premier League team. [Blog post].
  • AIrgentina is a model for predicting the results for the 2022 World Cup using a framework based on the team-level model used in the AIrsenal. We modified AIrsenal to make it more suited to predicting international results, and our resulting AIrgentina model came 6th in @Futbolmetrix1's WC2022 Sophisticated Prediction Contest outperforming models developed by FiveThirtyEight, Opta and Betfair! [Blog post].
    • Our model also got picked up by The New Scientist.
• Recommendation Systems for Podcast Discovery (ATI Data Study Group, April 2021)
  • Participated in a project with Entale to develop podcast recommendation systems, where I primarily worked on developing a topic model to recommend new podcasts based on the similarity to the topics that have been of interest to a listener previously.
  • Allowed me to gain some general experience working with natural language processing methods, collaborative filtering, text mining, clustering algorithms, dimension reduction techniques and recommender systems using Python. [Report].
• Bayesian Sports Modelling
  • In my masters project, I investigated the applicability of Bayesian hierarchical models for predicting the outcome of football matches. We were able to develop models that achieved a greater prediction accuracy than existing models in the literature.
  • R and Stan were used to implement various models; the code and report can be found here.

Selected Publications (Google Scholar)

• Daub, E. G. et al. Technical overview and architecture of the FastNet Machine Learning weather prediction model, version 1.0. 2025. [Report].
• Dunston, T. et al. FastNet: Improving the physical consistency of machine-learning weather prediction models through loss function design. 2025. [Pre-print].
• Chan, R.S.Y., Nanni, F., Lazauskas, T., Wood, R., Yong, P., Tarassenko, L., Girolami, M., Geddes, J., Duncan, A. Retrieval-augmented reasoning with lean language models. 2025. The Alan Turing Institute Technical Reports. [Report].
• Bowler, E., Byrne, J., Leclerc, L.M., Roberto-Charron, A., Rogers, M.S.J., Cavanagh, R.D., Harasimo, J., Lancaster, M.L., Chan, R.S.Y., Strickson, O., Wilkinson, J., Downie, R., Hosking, J.S. and Andersson, T.R. AI sea ice forecasts for Arctic conservation: A case study predicting the timing of caribou sea ice migrations. 2025. Ecological Solutions and Evidence. [Paper].
• Chan, R.S.Y., Nanni, F., Brown, E., Chapman, E., Williams, A.R., Bright, J., Gabasova, E. Prompto: An open source library for asynchronous querying of LLM endpoints. 2025. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies. [Paper].
• Williams, A.R., Burke-Moore, L., Chan, R.S.Y., Enock, F.E., Nanni, F., Sippy, T., Chung, Y.L., Gabasova, E., Hackenburg, K., Bright, J. Large language models can consistently generate high-quality content for election disinformation operations. 2025. PLoS ONE. [Paper].
• Alam, N., Kanjula, K.R., Guthikonda, S., Chung, T., Vegesna, B.K.S., Das, A., Susevski, A., Chan, R.S.Y., Uddin, S.M., Islam, S.B., Santhosh, R., Sharma, D., Liu, C., Chaturvedi, I., Winata, G.I., Mukherjee, S., Aji, A.F. 2025. In CVPR 2025 Workshop Vision Language Models For All. [Paper].
• Tseriotou, T., Chan, R.S.Y., Tsakalidis, A., Bilal, I.M., Kochkina, E., Lyons, T., Liakata, M. Sig-Networks Toolkit: Signature Networks for Longitudinal Language Modelling. 2024. In Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations. [Paper].
• Mougan, C., Plant, R., Teng, C., Bazzi, M., Ejea, A.C., Chan, R.S.Y., Jasin, D.S., Stoffel, M., Whitaker, K.J. and Manser, J. 2023. How to Data in Datathons. In 37th Conference on Neural Information Processing Systems (NeurIPS 2023) Track on Datasets and Benchmark. [Paper].
• Chan, R.S.Y., Johansen, A.M., Pollock, M., and Roberts, G.O. 2023. Divide-and-Conquer Fusion. The Journal of Machine Learning Research. [Paper].

Talks, Conferences and Awards

Awards

C4AI Expedition Aya 2024: Most Promising Award for Maya: Multimodal Aya
The Alan Turing Institute Doctoral Studentship (2018-2022)
The Royal Statistical Society Prize (2018)
The Top 10 Scholarship, awarded to the top 10 undergraduate students at the School of Mathematics, University of Leeds (2015, 2016, 2017)
Summer Vacation Bursary Scheme, awarded to undertake a research project with the School of Mathematics, University of Leeds (2016, 2017)


Presentations

Robots in Disguise: Mechanistic Interpretability III - Sparse Autoencoders. 20/11/24.
Robots in Disguise: Mechanistic Interpretability II - Circuits and Superposition. 02/10/24.
Robots in Disguise: Mechanistic Interpretability I - Introduction. 23/09/24.
Robots in Disguise: An overview of Llama 3.1. 12/08/24.
Robots in Disguise: Vision Transformers. 07/08/23.
Robots in Disguise: GPT. 24/07/23.
Robots in Disguise: BERT. 07/07/23.
Greek Stochastics Nu [Naxos]. 07/07/23.
Robots in Disguise: Transformer Encoder and Decoder models. 26/06/23.
Robots in Disguise: Sequence-to-sequence models: Part II - Encoder-Decoder models. 03/05/23.
Robots in Disguise: Sequence-to-sequence models: Part I - RNNs/LSTMs. 17/04/23.
AIUK Demo (IceNet) [London]. 2023. 21/03/23.
Lunchtime Tech Talk [The Alan Turing Institute]. 25/10/22.
2022 World Meeting of the International Society for Bayesian Analysis [Montreal]. 28/06/22.
Data Study Group (April 2021) Entale [The Alan Turing Institute]. 29/04/21.
Young Researchers' Meetings (YRM) [University of Warwick]. 09/02/21.
Statistics Group Seminar [Newcastle University]. 10/07/20.
RSS Discussion Meeting for "Quasi-stationary Monte Carlo methods and the ScaLE Algorithm" [The Royal Statistical Society]. 24/06/20. [Video].
The Student Seminar Series [The Alan Turing Institute]. 21/04/20.
University of Warwick Departmental Conference [Gregynog]. 24/03/20.
Algorithms & Computationally Intensive Inference Seminar [University of Warwick]. 08/03/19.
Student Seminar [University of Leeds]. 22/05/18. [Handout].

Posters

2022 World Meeting of the International Society for Bayesian Analysis [Montreal]. 28/06/22.
Bayes at CIRM 2021 [Marseille]. 25/10/21.
2021 World Meeting of the International Society for Bayesian Analysis [Online]. 28/06/21.
The Turing Research Showcase [Online]. 24/09/20. [Video].
Greek Stochastics Lambda [Corfu]. 28/08/19.
O'Bayes 2019 [University of Warwick]. 29/06/19.

Miscellaneous

Cycling from Manchester to London for Ambitious for Autism
Thesis: Monte Carlo methods for combining sample approximations of distributions; Examined by Professor Nicolas Chopin and Dr. Krzysztof Łatuszyński
Advent of Code 2022 with Julia
R package to implement Fusion methodologies (for unifying distributed analyses): DCFusion
R package to simulate layered Brownian bridges: layeredBB
Advent of Code 2021 with R
Masters Thesis: Bayesian Sports Modelling
Undergraduate project: Automatic Puzzle Solving
Fundraising for FareShare: Cycling challenge