One of the funding schemes of the NRDI Office is specifically designed to enable Hungarian students studying in UK universities to join internship programmes in Hungary. Grantees have the opportunity to return to Hungary for a few months and participate in the scientific work of innovative domestic businesses and research institutions to gain an insight into Hungarian career opportunities and the incentive system of research funding. This year’s winners are going to participate in research projects in solid-state physics, artificial intelligence and robot programming, to mention just a few.
In the world of science, the cross-border transfer of knowledge and the free international flow of discoveries and ideas have been key principles for centuries. This tradition is represented by Hungarian students studying abroad, for example in the United Kingdom. However, for Hungary it is a primary concern that young Hungarian researchers who have gained expertise and reputation abroad later return to their country and continue their research career at home. For this, however, they need to learn about research career opportunities and research funding options available in Hungary.
This aim is promoted by the Funding for Hungarian internship programmes for Hungarian university students studying in the UK (2018-2.1.1-UK_GYAK) call, the latest funding decisions of which have been recently announced by the NRDI Office. The grant enables talented young Hungarian students to join summer internships at innovative businesses and accredited research groups at higher education institutions and research centres in Hungary. We have asked the host institutions of some grantees about their expectations toward the interns.
Artificial Intelligence in breast cancer treatment
Despite his young age, Márk Di Giovanni, grantee hosted by the Physics of Complex Systems. Department at Eötvös Loránd University (ELTE), has already made a reputation in the scientific community. As we learnt from prof. István Csabai, Márk studies Mathematics at Cambridge University but his talent was already apparent in high school. He twice won the National Secondary School Academic Competition and was regularly among the top performers in the calls of the Matematikai Lapok mathematical journal.
‘I do not exaggerate if I say that Márk is one of the most talented members of his generation. He happened to be thinking about a topic similar to ours: the application of Artificial Intelligence in medicine,’ said István Csabai, professor at ELTE, in response to our question. ‘So, he will join our research project on machine learning based medical diagnostics, which is also the theme of PhD student Dezső Ribli. Dezső has developed a programme which has recently earned him the second place out of 1200 registered competitors in the first round of The Digital Mammography DREAM Challenge, a large international scientific data analysis competition. Then in the second round, a further developed version of his method proved to be the most effective solution by far.’
The present research was enabled by enormous recent progress in the application of artificial neural networks. Imaging methods have become drastically more accurate than their predecessors, producing results that are now comparable to human accuracy. With algorithms becoming increasingly precise and computers increasingly fast, scientists are now able to apply these methods to scientific problems. Large internet companies are also among the frontrunners in imaging development, though they are mostly focused on everyday online images. But image processing has a much more critical area of application.
‘X-ray images made for breast cancer diagnostics hardly make any sense to the layman’s eye, and even radiologists learn for years how to evaluate them. The evaluation of the growing amount of data produced by more and more, increasingly productive examination methods can be quite burdensome for doctors, and there is also a shortage of suitably qualified professionals,’ István Csabai argues. ‘Mammography images, for instance, are examined by two radiologists to filter out any errors. We hope that one of them can be replaced by a machine learning software, enabling doctors to focus on more complicated cases rather than routine analyses. Artificial Intelligence, however, needs special teaching to recognise the signs of cancer. This is where interns will be involved.’
Base for a long-term internship programme
The highest number of successful project proposals were submitted by the Wigner Research Centre for Physics of the Hungarian Academy of Sciences (HAS) who are going to host five interns for the summer. Their success is partly due to the fact that they had already planned to launch an internship programme before the announcement of the call for proposals, so the funding opportunity came just in time.
‘Péter Lévai, Director of HAS Wigner Research Centre took part in a roadshow popularising Hungarian research opportunities in Cambridge, where students showed great interest in joining the work of the institution,’ Gergő Orbán, research fellow at HAS Wigner Research Centre begins. ‘So, it’s no wonder that four of the interns are going to come from Cambridge and only one from Birmingham. They will participate in diverse research projects, as our goal was to offer them a wide scope of opportunities. Research topics cover solid-state physics, quantum information theory and even theoretical physics.’
The Research Centre considers the internship programme as an investment. The aim is to put Hungarian research opportunities into perspective for students studying abroad. Professional practice introduces interns not only to domestic research projects but also to domestic research funding schemes.
‘Besides providing professional experience for the students, we also want to show them the various career development instruments available here. My experience shows that the funding opportunities available for a researcher in Hungary are much less visible from abroad,’ says Gergő Orbán. ‘Students are evidently interested in this; they come here to work on their topics but they also want to get information on domestic opportunities. They will not necessarily start PhD studies in Hungary immediately after graduating but they may have easier access to the road back to Hungary after obtaining the doctoral degree.
The research fellow believes that the internship programmes will exert their effects in the long run. If the students gain useful research experience here, it will open the way directly for them and for the entire student community to a predictable career path. It is our belief that we are laying the foundations of a long-term internship programme.
Robots thinking independently
The European Centre of Excellence in Production Informatics and Control (EPIC) was recently established at the HAS Institute for Computer Science and Control (SZTAKI). The intern, who is a mechatronics (mechanical engineering and electronics) student at Cambridge University, is going to take part in a robot development project at the centre.
‘We make robots “smart” with various sensors and with IoT integration we teach them to perform tasks which so far only humans have been able to perform,’ says Gábor Erdős, fellow researcher at HAS SZTAKI. ‘We have integrated camera, load sensor and laser interferometers into the controller of the robots. The latter units are small computers fitted on the robots.’
Usually, robots do not process the captured data but forward it to a remote computer and receive the adequate response to a given reaction. This data processing method, however, makes the operation of robots slow and inefficient. So, it is going to be the intern’s task to develop a method enabling these calculations to take place within the robots.
‘During the summer, the intern will have to find out how to write software that would enable algorithms currently running on remote computers to run on the controllers,’ Gábor Erdős explains. ‘This will make information processing faster and the system more stable, making errors easier to detect. Eventually, such sensors empowered with on-the-spot data processing ability will be able to play a much more effective role in the Internet of Things.’
Is grade retention a good solution?
The HAS Centre for Economic and Regional Studies will also host a Hungarian intern who is about to start his master’s course in economics at Cambridge, one of the most prestigious universities in the United Kingdom. This time, it was the intern who contacted the host institution with the idea of application. The research topic for the summer practice was offered to him by the host institution.
‘The intern will join a new research project at our research institution which addresses the effects of grade retention on the retained student’s performance,’ said Zoltán Hermann, research fellow at the HAS Centre for Economic and Regional Studies. ‘The question is whether retained students perform better in the next year and what are their chances of dropping out.”
The research will mostly analyse the data of primary school students but 9th and 10th graders will also be involved. No new (e.g. questionnaire) survey will be conducted, but the project will rely on administrative data instead, mainly taken from the national competence survey. Test results achieved in the survey will be connected with other school data, including, of course, data on grade retention. This database will form the basis of statistical analyses.
Although this will be a pioneer project in Hungary, some data in international literature can be used to formulate hypotheses about the expected results.
‘Literature shows that grade retention has either no or even negative impact on the school performance of the retained student. So, we are very curious to see what the situation is in Hungary,’ Zoltán Hermann adds. ‘Poorer performance may be due to social stigmatisation or further reduced motivation which makes retained students to put even less energy and effort to studying, making them frustrated.’