Statistics and Data Science training to build research capacity to address societal problems in Mongolia

Mongolia is a developing country that is large in area but small in population. Since the fall of communism, Mongolia has experienced difficulty in developing both an effective civil service and a university system that generates the expertise to solve the problems of a modern society. Much information is gathered but institutions and stakeholders lack the experience to extract knowledge from this data and use it effectively in the public interest. We propose a programme of workshops that will train Mongolian academics, civil servants and other stakeholders in the use of statistics, machine learning and data science. The programme will be customised to work on some of the current most pressing Mongolian social challenges, using real data accumulated by Mongolian public institutions. This will ensure that the technical knowledge shared is not abstract but applied with the potential for immediate impact on Mongolian society.

The programme will have 4 stages.

In the first stage we will scope out the problems and assemble the data. We have the advantage of our prior collaborations with Mongolia which will already offers a head start.
We follow this in the second stage with a two week session of instruction in the capital of Mongolia, Ulaanbaatar. This will be given by expert academics with the support of well-trained doctoral students. The instruction will be highly interactive using modern teaching techniques to accommodate for the diversity in previous experience of participants as well as diversity in English language ability. Graduate students from the University of Bath will play a significant role in teaching and interacting with the Mongolians as they learn. The team has previous experience in delivering similar activities to civil service data scientists at the Paraguayan Ministry of Development. Best practice from this context will be integrated into the delivery in Mongolia.

In the third stage, four leading Mongolian stakeholders will visit Bath for a week to attend one of Bath's innovative Integrative Think Tanks (ITT). The latter is a unique interactive research training workshop that has been developed and already used multiple times for Bath's EPSRC Centre for Doctoral Training in Statistical Applied Mathematics (SAMBa). Such workshops are based around the principle of taking real world challenges in industry, or in this case, the public sector, and, through a tried and tested process of facilitation, distilling out well-formulated mathematical research questions that can move forward into an active and collaborative research projects.

The ITT in Bath will use as its starting challenges the specific Mongolian data and societal problems described earlier. This will also involve a larger number of Bath academics and (SAMBa) graduate students who will contribute to and learn from the aforementioned process of distilling out tangible mathematical questions. Having run nine ITTs, and with the accolade of generating manifold new research projects with accompanying funding support, the Department of Mathematical Science at Bath has a strong track record in the delivery of ITTs.

The final stage will be a week long visit back to Ulaanbaatar to further advance the research agenda on the Mongolian problems by feeding them back into an ITT attended by a wide spread of Mongolian stake-holders together with Bath academics and doctoral student.

Throughout, we will be in electronic communication with our Mongolian partners to advance the instruction and progress on research.

At the end of the programme, we will have trained a group of Mongolian academics and civil servants in statistics and data science. We will have advanced understanding on a range of Mongolian problems using their data. Knowledge will be transferred in that the academics will be able to pass it on to their students and the civil servants will have developed data-driven policy making skills.

There are a number of mechanisms by which create either pathways to impact or direct impact as follows.

1. The training packages will work with genuine Mongolian challenges and Mongolian accumulated data and so the step from learning to applicability and impact is already built in.

2. Academics will work directly with those Mongolian public agencies and NGOs which reporting and advise directly to the parliament and cabinet executive on the three intertwined socio-economic challenges (identified in this grant proposal) allowing for impact on national public policy.

3. More generally, Mongolia is a relatively small country in population (around 3M people) which makes the networking distance from citizen to the cabinet executive of government very small in general. This means there is genuine opportunity for training and research ideas to permeate into public policy at the highest level.

4. The principal application areas which will gain impact through the activities of the grant informing public policy will be:
- addressing the very serious problem of heavily toxic air pollution (Ulaanbaatar is the world's most polluted capital city)
- classifying in a quantitative way, the fine line between those who have modest but adequate resources and those who are living in poverty
- establishing data-informed policy which reduces the scope for corruption in economic policy.

5. More generally, there will be obvious impact on the transferable skill set of the Mongolian academics and government data-scientists which will be enhanced throughout the training programme delivered through this proposal.

6. There will also be impact on the large body of UK-based doctoral students involved in this bid, who will gain an exceptional experience and develop transferable skills in the use of data science for the development of policy with social or economic impact. This can play out e.g. into UK-based public service and/or industry.

Other Bath academics not named specifically in this proposal will participate in the workshops. Given the experience of previous ITTs stimulating new research projects, which are taken forward by both graduate students and academics, we would expect this may benefit their research in unforeseen ways and develop into fruitful collaborations. As well as academics from mathematical sciences, we will also ensure that the workshops are populated by academics from other departments with relevant expertise. A pertinent example in this respect is the inclusion of some academics from the Bath Institute for Policy Research and the Bath Institute for Mathematical Innovation.

We have a level of access to data in Mongolia and the potential to influence government policy which is difficult to achieve in more developed countries. There is great potential, particularly in collaboration with Institute for Policy Research at the University of Bath, to have real impact directly into some of the most serious Mongolian social challenges and learn lessons that can be used elsewhere, even in the UK.

We will openly share our experience with other UK academics who may have partnerships with other developing countries and may be interested in following a similar programme.