Department of Engineering, University of Cambridge
Supervisor: Prof Flavia Mancini
Application deadline: Friday, December 04, 2026
Funding: Competition Funded PhD Project (UK Students Only)
About the Project
AI_CDT_DecisionMaking
Details
Adaptive neurotechnology must make stimulation decisions from incomplete information about a changing nervous system. This PhD will develop interpretable machine learning methods to quantify uncertainty and guide interventions, with chronic pain as the main application. The project will use complementary neural recordings to develop approaches that generalise across experimental settings
The central question is: how can models of partially observed neural dynamics support reliable stimulation choices when intervention effects are uncertain? Different models can explain the same recordings while predicting different responses to stimulation. The project will investigate how to identify and resolve the uncertainties that matter for choosing an action.
First, the student will develop structured probabilistic models relating observed activity to latent neural states, measured inputs and unobserved drive. Comparisons between ongoing and stimulus-evoked activity will test how preceding neural state and stimulus history shape responses. Model complexity will be justified against interpretable baselines using held-out prediction and calibration. Recovery experiments will establish which properties can be inferred reliably; inferred inputs will not automatically be assigned a biological source.
Second, the project will develop sequential experimental design methods that select informative stimulation conditions from an expert-approved set. A particular focus will be disagreement between models that explain existing data similarly but recommend different actions. The hypothesis is that resolving such disagreement can improve decisions with fewer experiments than fixed protocols or methods that reduce uncertainty without considering its consequences for action.
Third, the student will evaluate adaptation under recording noise, missing observations, drift and variation between sessions or subjects. Experimentalists will specify permissible stimulation conditions, response limits and experimental costs, and scrutinise predicted responses and uncertainty before accepting or revising recommendations. Evaluation will assess predictive calibration, experimental efficiency, decision quality and constraint violations.
An initial audit will establish recording quality, simultaneous sampling, intervention coverage and metadata completeness. The primary action set and neural response endpoint will be selected with collaborators. Methods will be evaluated through controlled simulation and retrospective analysis, with prospective experiments where access and protocols permit. Neural response changes will not be equated with pain relief, and simulation results will not be treated as clinical safety validation.
Expected outputs include new methods for decision-relevant uncertainty, reproducible evaluation tools and an interpretable framework for adaptive experimental decisions. Chronic pain provides a focused test case, with broader methodological relevance to neurotechnology for Parkinson’s disease and psychiatric disorders. The four-year programme will accommodate the CDT’s substantial first-year training. Applicants should have strong quantitative and programming skills and an interest in developing machine learning through collaboration with experimental neuroscience.
Student Desirable Background
A strong quantitative background in machine learning, statistics, mathematics, computer science, engineering or computational neuroscience is desirable, alongside Python programming and knowledge of linear algebra, probability and statistical inference. Experience with machine learning and related areas such as time-series analysis or Bayesian methods is beneficial. Neuroscience experience is not essential. Applicants should enjoy interdisciplinary research, experimental data and developing reliable AI methods.
By submitting an application for a University of Cambridge project, you consent to your documents being shared with the relevant supervisory team at the partner institution. If you have questions about this process [contact details available in the full listing] before applying.
How to apply
Please apply under the University of Manchester application portal and select 'PhD in Artificial Intelligence' (OAA Application form)
When applying, you must include the project’s full name, your supervisor’s name, funding status, academic history, and two referees’ contact details. Applications with missing documents will not be processed. Late or incomplete submissions will not be considered, and the university is not responsible for missed deadlines.
After you have applied you will be asked to upload the following supporting documents:
- Final Transcript and certificates of all awarded university level qualifications
- Interim Transcript of any university level qualifications in progress
- CV
- Supporting statement: A mandatory one- to two-page statement explaining your motivation for postgraduate research, why you chose Manchester, relevant academic or work experience, key research findings, and the techniques or skills you've gained. Applications will be put on hold if this is not provided.
- Contact details for two referees (Please ensure referee email addresses are official university or workplace accounts, as we may need to verify the references.)
- English Language Certificate (if applicable)
Equality, diversity and inclusion is fundamental to the success of The University of Manchester, and is at the heart of all of our activities. We know that diversity strengthens our research community, leading to enhanced research creativity, productivity and quality, and societal and economic impact. (Equality, diversity and inclusion | The University of Manchester)
We actively encourage applicants from diverse career paths and backgrounds and from all sections of the community, regardless of age, disability, ethnicity, gender, gender expression, sexual orientation and transgender status.
We also support applications from those returning from a career break or other roles. We are dedicated to supporting work-life balance and offer flexible working arrangements to accommodate individual needs. Our selection process is free from bias, and we are committed to ensuring fair and equal opportunities for all applicants.
We are dedicated to supporting work-life balance and offer flexible working arrangements to accommodate individual needs, including a part-time study option.
Join us and help grow the UK’s national capability in artificial intelligence, part of the UK’s modern industrial strategy. Everyone is welcome if you are passionate about uncertainty quantification, decision making with humans in the loop and decision making for machine learning systems and their applications in Science and Engineering, and looking to develop your research and innovation skills at the doctoral level to further your career in the AI sector.
Funding Notes
Fully funded AI UKRI CDT 4 year program; The Project will be based in University of Manchester.
The UKRI AI CDT in Decision Making for Complex Systems is part of the TechFirst offering stipends of £30,000 for students eligible for home fee status. The start date is September 2027.
The Project will be based in University of Cambridge.
References
Pandarinath, C. et al. (2018). Inferring single-trial neural population dynamics using sequential auto-encoders. Nature Methods, 15, 805–815. https://doi.org/10.1038/s[contact details available in the full listing]
Foundational work on learning latent dynamics from neural recordings.
Foster, A., Ivanova, D. R., Malik, I. & Rainforth, T. (2021). Deep Adaptive Design: Amortizing Sequential Bayesian Experimental Design. Proceedings of ICML, PMLR 139, 3384–3395. https://proceedings.mlr.press/v139/foster21a.html
A methodological foundation for selecting informative experiments adaptively.
Vahidi, P., Sani, O. G. & Shanechi, M. M. (2025). BRAID: Input-Driven Nonlinear Dynamical Modeling of Neural-Behavioral Data. ICLR 2025. Directly relevant to separating intrinsic neural dynamics from the effects of measured external inputs.
Sani, O. G., Pesaran, B. & Shanechi, M. M. (2024). Dissociative and prioritized modeling of behaviorally relevant neural dynamics using recurrent neural networks. Nature Neuroscience, 27, 2033–2045. Relevant to identifying neural dynamics that matter for a particular functional outcome.
Oehrn, C. R. et al. (2024). Chronic adaptive deep brain stimulation versus conventional stimulation in Parkinson’s disease: a blinded randomized feasibility trial. Nature Medicine, 30, 3345–3356. A clinical example of personalised adaptive stimulation, demonstrating feasibility in a small trial.
Steel, K. et al. (2026). Contribution of Wide Dynamic Range neuronal activation to somatosensory evoked potentials supports use as biomarker of spinal nociceptive processing. bioRxiv, preprint. The experimental foundation for the Bristol spinal recordings.
EPIONE — Effective Pain Interventions with Neural Engineering. EPSRC–MRC Research Programme. https://eng.ox.ac.uk/epione
Programme context linking neuroscience, engineering and adaptive interventions for chronic pain.
NOX LAB - https://www.noxlab.org/ (new website to be released soon)
