MCMC methods are a great tool for fitting data because they explore the whole parameter space and, more importantly, they are able to deliver uncertainties. Uncertainties are crucial because they show how reliable the fit is. When the fit looks reasonable and the uncertainties are not very high, you can claim that you were able to describe your data successfully. However, what happens when your fits do not look so good; either because the values are unrealistic in the context of the system you are studying, or because the uncertainties are too high for the fits to be reliable? Is it the data or the approach to fitting the data the cause of failing?
In this talk, I will talk about my personal experience using MCMC methods during the course of my PhD. Firstly, I will show through different examples how the physical interpretation of the fitted values and their uncertainties was crucial in solving problems in my research. Sometimes, failing to fit the data –especially due to high uncertainties– led me to new insights about system I was struggling to understand, even taking that project into a whole new direction. Secondly, I will talk about situations where I still struggle fit the data, and I will share my insights as to why.
Nowadays, the volume of science or engineering data has increased substantially, and a variety of models have been developed to help to understand the observations. Markov chain Monte Carlo (MCMC) has been established as the standard procedure of inferring these model parameters subject to the available data in a Bayesian framework. Real systems such as interacting galaxies require complex models and these models are computationally prohibitive. The goal of this project is to provide a flexible platform for connecting a range of efficient algorithms for any user-defined circumstances. It will also serve as a testbed for assessing new state-of-the-art model-fitting algorithms.
The most commonly used MCMC methods are variants of the Metropolis-Hastings (MH) algorithm. At the beginning of this project and in this article, the standard MH-MCMC algorithm together with affine-invariant ensemble MCMC, which has dominated astronomical analysis over the past few decades, has been tested to reveal the performance of each sampler for the problems with known solutions. The Hamiltonian Monte Carlo algorithm was also tested and it shows in which circumstance that it outperforms the other two.
In my talk, I will start by introducing the basics of Markov Chain Monte Carlo (MCMC) methods. I will start with the Metropolis and Gibbs samplers, and the proceed to the Hamiltonian Monte Carlo sampler. A key focus will be to describe the domains of applicability of each of these sampling methods and the difficulties they encounter when applied. I will then describe strategies to overcome some of the difficulties encountered with basic versions of these methods. I will also touch on output diagnostics, and determining when a sampler is working as desired. Finally, I will consider the case of sampling where the likelihood of a model is expensive to compute and how MCMC can be used in this situation. The latter case may be of interest in astronomy applications.
Various multiobjective optimization algorithms have been proposed with a common assumption that the evaluation of each objective function takes the same period of time. Little attention has been paid to more general and realistic optimization scenarios where different objectives are evaluated by different computer simulations or physical experiments with different time complexities (latencies) and only a very limited number of function evaluations is allowed for the slow objective. In this work, we investigate benchmark scenarios with two objectives. We propose a transfer learning scheme within a surrogate-assisted evolutionary algorithm framework to augment the training data for the surrogate for the slow objective function by transferring knowledge from the fast one. Specifically, a hybrid domain adaptation method aligning the second-order statistics and marginal distributions across domains is introduced to generate promising samples in the decision space according to the search experience of the fast one. A Gaussian process model based co-training method is adopted to predict the value of the slow objective and those having a high confidence level are selected as the augmented synthetic training data, thereby enhancing the approximation quality of the surrogate of the slow objective. Our experimental results demonstrate that the proposed algorithm outperforms existing surrogate and non-surrogate-assisted delay-handling methods on a range of bi-objective optimization problems. The approach is also more robust to varying levels of latency and correlation between the objectives.
Slice Sampling has emerged as a powerful Markov Chain Monte Carlo algorithm that adapts to the characteristics of the target distribution with minimal hand-tuning. However, Slice Sampling’s performance is highly sensitive to the user-specified initial length scale hyperparameter and the method generally struggles with poorly scaled or strongly correlated distributions. To this end, we introduce Ensemble Slice Sampling (ESS) and its Python implementation, zeus, a new class of algorithms that bypasses such difficulties by adaptively tuning the initial length scale and utilising an ensemble of parallel walkers in order to efficiently handle strong correlations between parameters. These affine-invariant algorithms are trivial to construct, require no hand-tuning, and can easily be implemented in parallel computing environments. Empirical tests show that Ensemble Slice Sampling can improve efficiency by more than an order of magnitude compared to conventional MCMC methods on a broad range of highly correlated target distributions. In cases of strongly multimodal target distributions, Ensemble Slice Sampling can sample efficiently even in high dimensions. We argue that the parallel, black-box and gradient-free nature of the method renders it ideal for use in scientific fields such as physics, astrophysics and cosmology which are dominated by a wide variety of computationally expensive and non-differentiable models.
The improvement of energy efficiency of existing buildings is key for meeting 2030 and 2050 energy and CO2 emission targets. Thus, building simulation tools play a crucial role in evaluating the performance of energy retrofit actions, not only at present, but also under future climate scenarios.
A Bayesian calibration approach, combined with sensitivity analysis, is applied to reduce the discrepancies between measured and simulated hourly indoor air temperatures. Calibration is applied to a test cell case study developed using the EnergyPlus building simulation software. Several scenarios are evaluated to determine how different variables may impact the calibration process: orientations, activation of mechanical ventilation, different blind aperture levels, etc. Uncertainties associated with model inputs (fixed parameters in the energy model), model discrepancies due to physical limitations of the building energy model (simplifications when compared to the real performance of the building), errors in field observations and noisy measurements were also accounted for.
Even though uncalibrated models were within the uncertainty ranges specified by the ASHARE Guidelines, pre-calibration simulation outputs over-predicted measurements up to 3.2 ºC. After calibration, the average maximum temperature difference was reduced to 0.68 ºC, improving the results by almost 80%. Thus, these techniques are proven to improve the level of agreement between on-site measurements and simulated outputs. Besides, the implementation of this methodology is useful for calibrating and validating indoor hourly temperatures and, consequently, provide adequate results for thermal comfort assessment.
With a few lines of Python code, it is possible to train a neural network over a data-set and then use it to make extrapolations.
These techniques are routinely used in particle physics and condensed matter and only recently some pioneering work has been done to apply them to the nuclear physics case.
Given the incredible potential of these techniques, is it still necessary to invest time to build complex nuclear models to do the same thing?
Why not directly use machine learning algorithms to analyse the data?
In my talk, I will present some applications of machine learning techniques to the case of nuclear masses: by using either neural networks [1] or Gaussian Process Emulators [2] I will show how to use these algorithms to reproduce this particular observable. In particular I will consider the case of a neural network to reproduce nuclear masses without any underlying model and with a model to improve performances.
This procedure may be very helpful: in the short term, it will help us detect possible trends in the data and eventually perform reliable predictions in nearby regions of the nuclear chart; in the long term, by interpreting the algorithms, we may learn what is the missing physics in the nuclear model we currently use.
[1] Pastore, A., & Carnini, M. (2021). Extrapolating from neural network models: a cautionary tale. Journal of Physics G: Nuclear and Particle Physics
[2]Shelley, M., & Pastore, A. (2021). A new mass model for nuclear astrophysics: crossing 200 keV accuracy. Universe, 7(5), 131
Many networks such as communication, social media, covert and criminal networks have event-driven dynamics where the intensity rate of the events changes according to the occurrences of events in the network. In particular, events that occurred in a node of the network could increase the intensity of other nodes depending on their causal relationship. Thus, it is of interest to use data to uncover the influence network in which the edges represent the directional influence between nodes. An event-driven dynamic on a network can be modelled by a multi-dimensional Hawkes process driven by count data. In this setting, the influence structure of the network is then parameterized by the Hawke process. Understanding the uncertainty of the network constructed from the data is also important. This talks will discuss how we may build an ensemble of networks to reflect uncertainty. The outcome will facilitate downstream uncertainty analysis for network applications such as node classifications, link prediction and rare-event detection.
Quantifying model uncertainty and performing model selection within a Bayesian framework is becoming an ever-larger part of scientific analysis both within and outside of astronomy. I will present a brief introduction to Nested Sampling, a complementary framework to Markov Chain Monte Carlo approaches that is designed to estimate marginal likelihoods (i.e. Bayesian evidences) and posterior distributions, outline some of their pros and cons, and briefly discuss more recent extensions such as Dynamic Nested Sampling. I will also briefly highlight `dynesty`, an open-source Python package designed to make it easy for researchers to applying Nested Sampling approaches to various “black box” likelihoods present in their work.
Computational fluid dynamics (CFD) is a simulation technique widely used in chemical and process engineering applications. However, computation has become a bottleneck when calibration of CFD models with experimental data (also known as model parameter estimation) is needed. In this research, the kriging meta-modelling approach (also termed Gaussian process) was coupled with expected improvement (EI) to address this challenge. A new EI measure was developed for the sum of squared errors (SSE) which conforms to a generalised chi-square distribution and hence existing normal distribution-based EI measures are not applicable. The new EI measure is to suggest the CFD model parameter to simulate with, hence minimising SSE and improving match between simulation and
experiments. The usefulness of the developed method was demonstrated through a case study of a single-phase flow in both a straight-type and a convergent-divergent-type annular jet pump, where a single model parameter was calibrated with experimental data. This talk is based on a journal article we previously published in the AIChE Journal.
Since 2005 the UK higher education system has hosted a steadily increasing number of European students from outside the UK. EU students hold a widespread reputation of being capable and driven, and these qualities have made them desirable to UK universities. While their participation varies between institutions, and between the hierarchical layers of the sector, they have become recognised as a vital contributor to the diversity of the student fabric on UK campuses. However, following the United Kingdom’s exit from the European Union (commonly referred to as Brexit) EU students will soon pay higher fees in the UK and lose access to the UK’s pay later tuition loans. Additionally they will be subjected to visa requirements and their post-study stay will be constrained by migration rules. Consequent to these changes, among others, it is anticipated that the amount of EU students opting to study in UK universities may decline by up to half of their pre-Brexit numbers. These projections provide a window through which we can examine what the potential loss of European students would mean for institutions across the UK. To that end, this paper examines interviews conducted in 12 UK universities with 127 participants pre-Brexit, predominantly senior executives and members of academic leadership. The analysis uncovered a number of institutional representations of EU students that arose in response to Brexit, most often concerning: student numbers and income; diversity; and quality. Representations varied geographically across the different nations of the UK, largely due to the differences in funding regimes unique to each but also institutional hierarchies within a stratified higher education system. The specificities of institutional representations within each nation highlighted the differential impacts the loss of EU students may have on universities across the sector, with notable implications for: future recruitment strategies; intra- and international competition; the breadth and nature of subject and study programme offerings; the creation and maintenance of collaboration networks; and interactions between students, funding, and research.
The educational migration of international tertiary students is a continuing worldwide trend. However, not everyone has sufficient resources for international education. Thus without state support, the cross-border mobility of students remains less than optimum. The state scholarships for outward student mobility pave the mobility way for students from lower-income groups.
In this study, as case studies, the state scholarships of Turkey and Chile were examined because of the differences in return obligation procedures that influence participation. While the Chilean scholarship programme obliges its recipients to return to Chile in 2 years after graduation and stay in Chile for a period depending on in which region the recipient lives, the Turkish programme appoints its recipients at pre-selected positions to work for the twofold study duration.
8 doctoral scholarship holders studying in the UK (4 for each group) were interviewed in 2019 to understand how the specific structure of a scholarship programme influenced their motivations to participate. After the data transcription, thematic analysis was applied, and the emerged themes as follows; quality overseas education, intercultural experience and (English) language acquisition. Since all participants are from the working class, receiving an international education was the primary motivation. Additionally, some Turkish participants stated their reason to apply for the scholarship as a “job guarantee”, which the Turkish scholarship programme offers at the selected universities and public institutions following graduation.
Albeit, most of the offered positions exist in the less-developed regions of Turkey. Thus the obligatory working duration hinders participation. The Chilean programme does not offer a job. However, the Chilean participants expect to get a decent job in Chile in their “preferred field” due to the international degree. Briefly, the participants are mainly inclined to participate in international scholarship programmes to advance their career opportunities in their homelands, where the youth unemployment rate is quite high.
The European Union (EU) Erasmus+ program is the most common student exchange schema in European Higher Education Area (EHEA). The program included member states of the European Union, members of the European Free Trade Association (EFTA) countries (Norway, Iceland, Liechtenstein), and candidate countries (the Republic of North Macedonia, Republic of Turkey, and the Republic of Serbia). The universities awarded with Erasmus+ program funds do not have any country restrictions to send or receive students since Erasmus+ is seen as a means of unification under the EHEA and the Bologna Process. However, over the years, specific mobility patterns emerged between countries that reciprocally exchange students based on the universities’ inclinations to sign mobility agreements predominantly with other universities in certain regions in Europe. These patterns portrait not a unified higher education area in Europe but a fractured one. It can be argued that these patterns heavily reflect the geopolitical positions and pragmatic preferences of individual countries involved in the Erasmus+ Program. These sub-regions and hubs in the EHEA are usually constrained by sectoral, practical, and historical positions and relations of universities involved in the Erasmus+ program. Hence, this research aims to analyze the network properties of the Erasmus+ student mobilities between these countries based on official statistics provided by the EU to determine a general account of the geopolitical hubs and sub-regions concerning the EHEA.
Outgoing student mobility in teacher training has, among other things, the purpose of developing culturally sensitive teachers. However, how likely students are to engage in various mobility schemes varies depending on where they are geographically located. International coordinators from teacher training colleges around Denmark tell the same story: The geography of the college highly influences the number of students engaging in outgoing mobility, because geography and the socio-demographic profile of the students is linked. The coordinators describe students in bigger cities, such as Copenhagen and Aarhus, as more adventurous and less tied in family and economic affairs, while students in the outskirts of Denmark are bound up by, among other things, family, property, and a generally more rooted lifestyle. The location of the teacher education, and accordingly the profile of the students, creates different challenges for the coordinators in relation to student mobility. To accommodate these challenges, the coordinators work on developing ‘mobility packages’ that make planning easier and secure access to mobility for all students. Furthermore, my findings indicate that the various coordinators employ different strategies in the process of attracting students to study abroad, such as dividing the students into yes-no-maybe groups depending on how likely they are to engage in mobility. This then influences which students they then spend time on guiding. I draw on data from my Ph.D.-project and I wish to present and discuss findings from interviews with ten international coordinators, and discuss possible solutions and how these aim at changing access to mobility.
English is the language of instruction in many universities around the world. Accordingly, English language skills, which youths of less privileged social backgrounds are often said to ‘lack’, plays an important role in the future aspirations of young people and to access and widening participation. In this paper, I argue that English is an important resource, which affects students’ mobility within and outside of their countries. Using interviews with 25 Colombian youths, some of which had finished secondary school and others were higher education students at the time of the fieldwork, my study explores the challenges and opportunities that Colombian students who plan to study abroad (including Europe) encounter in relation to English language. The findings indicate that the English capacity of the majority of the participants does not meet the English language requirements (mostly ILETS, TOEFL or GRE) of most universities with English as a medium of instruction. This ‘lack’ of English skills affects the social mobility and future aspirations of the students and, thus, their motivation to study abroad. Ultimately, the study reflects on the implications of English provision for social and academic mobility.
Transnational education (TNE) represents a lesser known aspects of the internationalization of higher education, whose volume and importance are growing.
If, on the one hand, TNE represents a way for universities to expand recruitment offshore; on the other hand, it also constitutes for students an alternative to student mobility to traditional destination countries.
The trickiest interrogations arising from TNE expansion are connected with the more general questions on whom offshore students are and why they decide to enroll this way. Though, data on this topic are still scarce.
This paper presents the results of a survey conducted among students enrolled in German TNE projects in several countries. The results reveal that TNE students cannot be considered a monolithic group; rather they have fairly heterogeneous motivations and attitudes. In particular, they show how, for some students, TNE enrollment seems to be a way to overcome the (perceived) limits of the higher education sector in the origin countries and that, in few cases, TNE was a way to avoid ‘involuntary student mobility’. In the majority of cases, however, it seems that TNE has been considered by the respondents as a “safer mobility”, a “trial run to mobility” and a way to acquire the resources needed to emigrate afterwards by those who wanted to go abroad but were not able to do it. Finally, some of respondents who did not desire to study abroad experienced a sort of ‘internationalization of aspirations’, leading them to desire to go abroad after graduation.
Until recently, internationalisation of higher education was largely considered an end in itself especially due to its impact on the UK economy. In the past few years, however, the conversation has changed considerably and with the challenges that Covid-19 restrictions have presented for the internationalisation based on students ‘mobility, the focus has diverted more on improving the quality of education and research as well as serving larger social goals. The focus should be on developing a culture of mutual support and strategic inclusion with partner Universities with potential development of dual purposing resource and mutual enrichment.
This contribution is aiming to investigate how does remote cooperative teaching, based on mutual enrichment across international ITE providers, support active participation of students in international activities?
Participants were undergraduate student on the ITE programme within Meduc year3 course: Health and Well-being Elective School of Education, University of Glasgow and Italian student teachers on the course “Scienze della formazione” at the Niccolo’ Cusano University, Rome.
Several sessions were organised and remotely delivered to both cohort of students with a careful blending of tutors’ expertise, focused on their cultural and language diversity; a carefully planned topic of high interest for both countries: Parental Engagement in pupil’s Education; and finally, a very well taught integration of the sessions’ content and task requirements into both Universities’ assessment agenda. Results showed a mutual enrichment and active participation which went beyond any expectations with elements of e-networking and overcome of language, communication and even possible stereotype barriers.
Mexico is the main Latin American country sending students abroad for international education. In 2018, there were 34,000 Mexican students enrolled in Higher Education Institutions (HEIs) outside their country of origin. From those, over 12,000 thousand (36%) studied in a European country. One of the most longstanding sources of funding for International Student Mobility (ISM) in Mexico has been the National Science and Technology Council (CONACYT) scholarship programme. This has funded graduate ISM since 1970 for the training and consolidation of human resources. In 2018, 58% of awardees studied in Europe. Some debates suggest these types of scholarships mainly benefit students with privileged backgrounds and reproduce inequalities in HE. This paper examines the socio-demographic profiles of former CONACYT scholarship doctoral awardees along with their different previous HE trajectories and decisions to study abroad.
This paper is drawn from my doctoral research using transformative learning theory concepts and the capabilities approach, exploring the transformative nature of ISM associated with individual meanings of the mobility experiences, capabilities developed, and implications for social change. In this paper, I present findings of the socio-demographic data collected through a cross-sectional survey and qualitative data from in depth semi-structured interviews. The findings show significant participation of students from less privileged backgrounds and a complex mix of drivers for outward mobility linked to previous educational opportunities and future life aspirations. This paper brings insights from a human development approach, showing how these scholarships are instrumental in increasing ISM opportunities, contributing to social mobility and facilitating career development.
European Commission along with its member states are targeting to attract and retain an increasing number of international students. Much of the earlier focus has been in what happens within the higher education context or in the public policy sphere while less attention is targeted on the communicative practices and how they (re)produce and position international students in the society. Communication forms the basis of social practices and the creation and continuation of social institutions like society (Voss & Lorenz, 2016) and higher education institutions. By taking the ongoing COVID19-pandemic as an example, we analyse the crisis communication targeting higher education students.
The aim of the paper is to analyse and compare actors, channels and content targeted at the international HE students in Finland during the pandemic time. The focus is particularly on the communicative practices of the national level actors. The main research question is how the communicative practices targeting higher education students differ between Finnish and international students. Using communication inequality (Ramandhan & Viswanath 2008, Viswanath 2006) as a theoretical framework, we specifically focus on the linguistic choices to differentiate these two student bodies. The data consist of ministerial documents and news pieces that are digitally provided at websites and social media sites. In a preliminary observation, we find some discrepancies among the actors in terms of the information provided and channels used supporting Palttala et al.’s (2012) observation on governments’ failing to communicate all different groups of people during a crisis.
This presentation will summarise the preliminary findings of our study which aims to document and analyse the experiences and well-being of Chinese and Chinese-presenting international students in the UK. The important contextual factors of this study include (1) the rise in hate crimes against East Asian communities in the early part of the COVID pandemic and (2) the continued increase in the number of Chinese students, who comprise the largest group of international students in the UK. Earlier research suggested that concerns around safety – both in relation to the COVID pandemic and racist incidents could affect student mobility from China.
Our study draws together data from freedom of information (FOI) requests, an online survey and semi-structured interviews to investigate (1) perceptions and experiences of aggression and (2) students informational practices during the pandemic. The latter refers to how students may have communicated their experiences of aggression and with whom. We employ insights from media studies, law and criminology, cultural sociology, and sociological studies of ‘resilience’ and social capitals to interpret our data. Our findings are intended to establish an evidence base to enable university staff and local communities to improve strategies and practices of safety and well-being of these students.