Gatsby Computational Neuroscience Unit MPhil/PhD

London, Bloomsbury

Students at the Gatsby Unit study toward a PhD in Computational and Theoretical Neuroscience and Machine Learning. You will network extensively with the Sainsbury Wellcome Centre and other UCL experimental groups. Gatsby is part of the Centre for Computational Statistics and Machine Learning, together with UCL Computer Science and Statistical Science. Approximately 90% of alumni have secured academic or industry positions.

UK students International students
Study mode
Full-time
UK tuition fees (2022/23)
£5,690
Overseas tuition fees (2022/23)
£27,750
Duration
4 calendar years
Programme starts
September 2022
Applications accepted
All applicants: 01 Oct – 15 Nov 2021

Applications closed

Entry requirements

Applicants must have a strong analytical background, a keen interest in neuroscience or machine learning and a relevant first degree at a minimum of upper second-class UK Bachelor's level or an overseas equivalent, for example in Computer Science, Engineering, Mathematics, Neuroscience, Physics, Psychology or Statistics. Students seeking to combine work in neuroscience and machine learning are particularly encouraged to apply.

If your education has not been conducted in the English language, you will be expected to demonstrate evidence of an adequate level of English proficiency.

The English language level for this programme is: Good

UCL Pre-Master's and Pre-sessional English courses are for international students who are aiming to study for a postgraduate degree at UCL. The courses will develop your academic English and academic skills required to succeed at postgraduate level. International Preparation Courses

Further information can be found on our English language requirements page.

If you are intending to apply for a time-limited visa to complete your UCL studies (e.g., Student visa, Skilled worker visa, PBS dependant visa etc.) you may be required to obtain ATAS clearance. This will be confirmed to you if you obtain an offer of a place.

Equivalent qualifications

Country-specific information, including details of when UCL representatives are visiting your part of the world, can be obtained from the International Students website.

International applicants can find out the equivalent qualification for their country by selecting from the list below. Please note that the equivalency will correspond to the broad UK degree classification stated on this page (e.g. upper second-class). Where a specific overall percentage is required in the UK qualification, the international equivalency will be higher than that stated below. Please contact Graduate Admissions should you require further advice.

About this degree

The Gatsby Unit is a world-class centre for Computational and Theoretical Neuroscience and Machine Learning. Our research seeks to understand the principles of learning, perception and action in brains and machines by developing mathematical algorithms. We provide a unique opportunity for a critical mass of theoreticians to interact closely and network extensively with each other and with other research groups across UCL, in particular the Sainsbury Wellcome Centre, and the Centre for Computational Statistics and Machine Learning. First year teaching is supplemented with weekly research talks, journal clubs and reading groups, external seminar programmes and participation in international conferences. As a student you will be supported by all members of academic staff, not just your immediate supervisors.

Who this course is for

What this course will give you

The Gatsby PhD programme was the first ever to combine theoretical neuroscience and machine learning within the same space. Our mathematical approach for developing novel algorithms and tools to understand learning, perception and action in brain and machines is unique. The most successful leaders in the fields of Computational Neuroscience and Machine Learning have studied or worked in the Unit, and the Unit has a reputation for offering world-class training. 

Our degree offers unique first year teaching in Systems and Theoretical Neuroscience (joint with SWC), Unsupervised Learning, Kernels, Theoretical Neuroscience and Advanced Methods in Machine Learning, as well as optional Deep Learning courses. In your second year you will have the opportunity to do a 2-month rotation in a lab of your choice, either at UCL or external. This rotation should be on a research area not related to your PhD topic. Our course also offers a variety of career development training. 

Since 2016, we are based in a new, state-of-the-art research facility in the heart of London and UCL. The building also houses the Sainsbury Wellcome Centre for Neural Circuits, and we interact closely with them. Our building offers two large seminar rooms used for teaching and other events, common areas for social interaction or quiet study, state-of-the-art IT facilities, kitchen facilities, and a large on-site brasserie offering breakfast, lunch, coffee and snacks. There are also two large outdoor areas and an on-site bicycle rack. 

The surrounding area is vibrant and offers a variety of shops, restaurants, cafes and bars, and museums. 

The foundation of your career

Our students receive exceptional world-class training through our intense, rigorous and diverse research programme. As a result, the majority are highly employable. Students will develop interdisciplinarity, and will be expected to collaborate widely through team work. You will develop your communication skills by giving regular internal talks and presentations. Critical thinking is encouraged through weekly journal clubs and lab meetings. You will be expected to teach during the second year of the programme and act as a mentor to junior PhDs. Participation in outreach and engagement is also encouraged. 

Employability

Most of our graduates have continued within the Computational Neuroscience and Machine Learning fields. 63% of PhD alumni have secured academic positions in top Institutions such as University of Cambridge, University of Oxford, Edinburgh University, MPI for Intelligent Systems, Columbia, Princeton, Caltech, École Normale Supérieure Paris and HHMI Janelia Research Campus. 25% of PhD alumni have gone on to work in organisations including Google DeepMind, Amazon, Facebook, Samsung, Babylon Health, and others. 

Networking

There are significant networking opportunities both within and external to the Gatsby Unit. Within the Unit there are daily teas where members take turns to present a 15-minute talk on an interesting and relevant topic to other members of the unit. Our weekly external seminar provides an opportunity to meet eminent speakers and external attendees. 

Based in the same building as the Sainsbury Wellcome Centre, you will network extensively with experimentalists, attend their seminars and weekly joint social events. Through CSML you will network with peers in Computer Science and Statistics. All students are encouraged to attend relevant conferences.

Teaching and learning

Approximate Inference and Learning in Probabilistic Models (COMP0085)

Probabilistic and Unsupervised Learning (COMP0086)

Advanced topics in Machine Learning (COMP0083)

Students are required to complete coursework for all first year courses. There are three examinations in the first year of study.

Typical contact hours are 9 - 5 pm.

Research areas and structure

  • Analysis of neural data
  • Neural dynamics
  • Neural plasticity
  • Perceptual processing of auditory and visual input
  • Neural population coding
  • Kernel methods
  • Bayesian statistics
  • Reinforcement learning
  • Statistical machine learning
  • Unsupervised learning
  • Network and relational data

Research environment

The Gatsby Computational Neuroscience Unit was established by the Gatsby Charitable Foundation at University College London (UCL) in July 1998 to provide a unique opportunity for a critical mass of theoreticians to interact closely with each other and with UCL’s other world-class research groups in neuroscience and related areas.  We were one of the first centres in the world to bring together the fields of theoretical neuroscience and machine learning, and our investigators have pioneered research into the mathematical underpinnings of learning, perception and action in natural and artificial systems.

In 2016 we moved from our original home in Queen Square to Fitzrovia in London's West End near the main UCL campus, to create a new collaborative partnership with the Sainsbury Wellcome Centre for Neural Circuits and Behaviour (SWC).  The SWC and GCNU work together closely, with shared research appointments, parallel PhD programmes, common day-to-day activities, and joint research projects that bring together theoretical and experimental neuroscience.  We are also part of the cross-faculty Centre for Computational Statistics and Machine Learning (CSML), and work with other UCL research groups in Neuroscience and Machine Learning

There are significant networking opportunities both within and external to the Gatsby Unit. Within the Unit there are daily tea talks where members take turns to prepare a ten-minute talk to other members of the unit. The unit also holds regular weekly seminars to which eminent speakers are invited; these are open to other members of the university and there is an opportunity to meet speakers and attendees after the talk. The Gatsby Unit schedule can be viewed online. Students also run their own reading groups and all students have the opportunity to attend relevant conferences and other international events.

The PhD programme lasts four years, including the first year of intensive instruction in techniques and research in theoretical and systems neuroscience and machine learning. Courses in the first year, taught in conjunction with colleagues from the SWC and CSML, provide a comprehensive introduction to theoretical and systems neuroscience and to machine learning; with further multidisciplinary training in other areas of neuroscience also available. Students are encouraged to work and interact closely with peers and faculty in the SWC and CSML throughout their PhD, providing a uniquely multidisciplinary research environment.

In the first term of Year 1, all students take the core courses in theoretical neuroscience (TN), systems neuroscience (taught with SWC) and machine learning (ML), after which they generally choose to concentrate on one of these fields. Students typically sit exams for TN and ML during the second term. In Year 2, students are expected to work on their thesis project. Students also do a 2-month rotation in a field that is not related to their PhD thesis. At the end of the second year, students write an upgrade report on their progress and any future plans, and give a presentation on this report to the rest of the unit. Having passed all the assessments, students then transfer from MPhil to PhD and are devoted to research for their remaining tenure. 

Throughout the training programme, students are immersed in a strong informal educational environment, comprising regular short informal talks, research reports, journal clubs and ad hoc reading groups; extensive seminar programs in the Unit itself and its UCL neighbours (including SWCICNFILPsychologyNeuroscience, Physiology and PharmacologyComputational Statistics and Machine Learning and Statistical Science); and attendance and participation in international conferences such as Cosyne, ICML, NeurIPS and Society for Neuroscience. 

Accessibility

Details of the accessibility of UCL buildings can be obtained from AccessAble accessable.co.uk. Further information can also be obtained from the UCL Student Support & Wellbeing team.


Fees and funding

Fees for this course

UK students International students
Fee description Full-time
Tuition fees (2022/23) £5,690
Tuition fees (2022/23) £27,750

The tuition fees shown are for the year indicated above. Fees for subsequent years may increase or otherwise vary. Where the programme is offered on a flexible/modular basis, fees are charged pro-rata to the appropriate full-time Master's fee taken in an academic session. Further information on fee status, fee increases and the fee schedule can be viewed on the UCL Students website: ucl.ac.uk/students/fees.

Additional costs

For more information on additional costs for prospective students please go to our estimated cost of essential expenditure at Accommodation and living costs.

Funding your studies

A number of competitive fully-funded studentships are available each year. These cover payment of full fees, provide a generous living stipend and covers travel and accommodation for conference attendance. Studentships are available to students of any nationality.

For a comprehensive list of the funding opportunities available at UCL, including funding relevant to your nationality, please visit the Scholarships and Funding website.

UCL Research Opportunity Scholarship (ROS)

Value: UK rate fees, a maintenance stipend, conference costs and professional development package (3 years)
Criteria Based on both academic merit and financial need
Eligibility: UK

Next steps

This PhD programme lasts for four years, including a first year of intensive instruction in techniques and research in theoretical neuroscience and machine learning. We only admit new students starting in September. Deadlines and start dates are usually dictated by funding arrangements so check with the department or academic unit (contact listed below) to see if you need to consider these in your application preparation. All applications for this programme must be made directly to the Gatsby Unit. Please refer to all information on the Gatsby Unit website for how to apply. The deadline listed on the Gatsby website is the correct deadline. Late applications will only be considered after all those received by the closing date. 

Please note that you may submit applications for a maximum of two graduate programmes in any application cycle.

UCL is regulated by the Office for Students.