The typical student in our CRT will follow a structured PhD training programming comprising research methods training, supervisor-initiated research-specific training, a CRT-organised training module in artificial intelligence methods, work placements, and international scientific research laboratories. In terms of specific CRT-organised training in artificial intelligence, we will focus on six thematic areas in which the co-applicant team and supervisor group have particular expertise. Theses are:
- Optimisation and Constraint Programming;
- Natural Language Processing;
- Machine Learning;
- Visual Media Processing; and
- Ethics of Data Analytics and Fair, Accountable, Transparent AI.
In our choice of areas, we are making no attempt at comprehensive coverage of the field of AI. Instead, we are selecting strands that (a) will relate to likely student PhD topics, and (b) afford scope for the students to acquire a broad range of relevant skills.
A typical student in our CRT will follow a structured PhD training programme that comprises four main elements: (i) Host-based research methods training; (ii) Supervisor-initiated research-specific training; (iii) CRT-organized training in Artificial Intelligence methods; and (iv) Work placements.
Host-based Research Methods Training. The host institutions expect all incoming PhD students to pass induction and research methods modules. Our CRT does not replicate these modules. They often include host-specific information and are useful in bringing students from different disciplines across the institution into contact.
Supervisor-initiated Research-specific Training. Supervisory teams within the in- stitutions currently work with students to choose additional modules to address knowledge-deficits and enhance existing skillsets. Supervisors will often also suggest that students attend short courses, online courses, summer schools, and so on.
CRT-organized Training in AI Methods. In addition to modules required by host institutions, and modules suggested by individual supervisors, all students who join the CRT will be expected to complete in their first-year a new ten-credit module in Artificial Intelligence Methods. This module is mandatory and cohort-based.
Placement Programme. A feature of this CRT is that students spend a significant amount of time working with partners through placements. Every student will spend three months during each of the first three years of their studies on placement at either and industry partner site or on secondment at an international research laboratory. We are modelling this approach on the US model of summer internship. Each year industry partners will be invited to interview students from each cohort for placement in their organisations. Students will work under the supervision of staff at the site of the industrial partner. Industry partners do not sponsor specific students for the duration of a cohort but, instead, have the flexibility to select from the CRT cohorts each year. While giving choice to industry partners, it also encourages students to think more long term in respect to the value that they can bring to industry projects.