Master of Science in e-Science

Purpose

The purpose of the Master of Science in e-Science is to create opportunities for learners in Computer Science, Statistics, Mathematics and Applied Mathematics, Engineering and other relevant disciplines to gain an interdisciplinary perspective of the emerging field of Data Science. It will prepare learners for professional knowledge and skills required for specific data science problem-solving in the workplace. This qualification will enhance learners' independent use of e-science tools to analyse and report on advanced data science problems. The purpose is to familiarise learners with communicating data science information effectively in a variety of inter-disciplinary scientific contexts. This qualification will enhance learners' decision-making capacity in applying advanced research methods in e-science across a range of scientific fields to produce high-quality research reports.

The qualification consists of two key components, namely coursework and mini-dissertation. The coursework consists of two compulsory modules, and four modules selected from the list of seven. These modules form the platform for the second component, namely the mini-dissertation.

Since this is an interdisciplinary qualification, the coursework cuts across specific broad-ranging fields that generate big data. For example, the coursework for this qualification will engage learners at a deep level in the following:

• Data Privacy and Ethics.
• Adaptive Computation and Machine Learning.
• Data Visualisation& Exploration.
• Large-scale computing systems and scientific qualification.
• Large-scale Optimisation for Data Science.
• Mathematical Foundations of Data Science.
• Special Topics in Data Science.
• Statistical Foundations of Data Science.

And to enable the mini-dissertation component, this qualification has chosen deep content engagement with a specific focus on research methods in data science and not the usual generic research methodology coursework.

As such, a compulsory 15 Credits module on research methods and Capstone Project in Data Science will scaffold the 90 Credits compulsory mini-dissertation in Data Science. This curriculum, purpose and qualification design will contribute to national and international priorities such as improving service delivery and developing a competitive knowledge economy. It will create opportunities for learners as professional researchers in academia and the public and private sectors. The curriculum will ensure that qualifying learners exit the qualification with the required knowledge, skills and competencies of a data scientist meeting NQF Level 9 descriptors.

Rationale

The school participated in the development of an MSc qualification within the National e-Science/e-Research Postgraduate Teaching and Training Platform (NEPTTP) to create opportunities for learners in Computer Science, Statistics, Physics and Applied Mathematics and any other related disciplines. The curriculum aims at preparing learners of this category for their participation in the growing e-Science necessitated by advances in cyberinfrastructure (computers/networks; data analytics/visualisation; data collection and storage, etc.). This will contribute to national and international priorities, such as improving service delivery and developing a competitive knowledge economy. It will create opportunities for learners as professional researchers in academia and the public and private sectors.

Learners seem to prefer the taught masters as opposed to research one because of the need to prepare learners for a new qualification that cuts across several knowledge fields. However, the qualification includes a research component on a 50/50 basis with a taught component in terms of credits. It would not be possible for individual departments to prepare candidates of this qualification adequately at the honours level and hence the need for a more or less self- contained qualification.

The qualification can be a useful transition across qualification levels if one is thinking of a career in academia but have not yet decided on an area of research and for those with less experience of self-study. In the case of research masters, necessarily, one would have to be self-motivated enough to drive one's studies at an acceptable pace.

Recognition of Prior Learning (RPL)

The institution's anchor the RPL policy upon life-long learning and the need to redress past inequities in higher education. Formal and informal learning experiences are measured, evaluated, and compared for purposes of establishing qualifications or level of learning possessed. This assessment will facilitate for enrolment of learners who do not meet the admission criteria or want to advance within the qualification's programme of learning.

Leaners who do not meet the admission requirements are required to carry out self-assessment to establish the gaps which before applying for RPL. If the identified deficiencies are those that could be filled by bridging modules, the learner could then proceed to register for those modules at the same time apply for RPL. The institution will advise the learners that do not qualify for RPL to register to do the necessary prerequisites before applying for admission. The institution will accept professionals employed as teachers or business practitioners in engineering or financial setups provided they have prior programming experience, sound undergraduate mathematical/statistical/physical/computer science/electronic engineering degree in conjunction with other sciences and an honours degree in any of the indicated disciplines.

Since the qualification is multidisciplinary, learners may apply RPL at entry or in the middle of the qualification. Therefore, there is a need for a system of credit accumulation and transfer (CAT) to facilitate smooth movement of learners between qualifications or institutions. Once prior learning has been measured and evaluated, by a subject specialist through the programme coordinator, a potential learner earns credits for learning acquired elsewhere.

A programme coordinator is responsible for guiding the learner through the application for RPL following the RPL Policy document. There are two important forms to be completed regarding application for RPL; RPL for access and RPL for credits. In cases where the evaluation of learning done elsewhere is contested, either by the coordinator or peers, the RPL assessment in the form of entry tests may be conducted before RPL is granted.

Entry Requirements

The minimum entry requirement for this qualification is

• Bachelor of Science Honours in Computer Science, NQF Level 8.

Or

• Bachelor of Science Honours in Mathematics, NQF Level 8.

Or

• Bachelor of Science Honours in Statistics, NQF Level 8.

Or

• Bachelor of Science Honours in Applied Mathematics, NQF Level 8.

Or

• Bachelor of Science Honours in Physics, NQF Level 8.

Or

• Bachelor of Science Honours in Electrical and Information Engineering, NQF Level 8.

This qualification consists of the following compulsory modules at National Qualifications Framework Level 9 totalling 225 Credits.

Compulsory Modules, Level 9, 225 Credits

• Research Methods and Capstone Project in Data Science, 15 Credits.
• Data Privacy and Ethics, 15 Credits.
• Adaptive Computation and Machine Learning, 15 Credits.
• Data Validation and Exploration, 15 Credits.
• Large Scale Computing Systems and Scientific Programming, 15 Credits.
• Large Scale Optimisation for Data Science, 15 Credits.
• Mathematical Foundations of Data Science, 15 Credits.
• Special Topics in Data Science, 15 Credits.
• Statistical Foundations of Data Science, 15 Credits.
• Mini-Dissertation: Data Science, 90 Credits.

1. Conduct high-level research in e-Science related projects derived from various disciplines including engineering or medical or biological or digital humanities or physical sciences or any other related fields.
2. Demonstrate a broad understanding of the field of study through embarking on research projects that cut across disciplines.
3. Participate in international research environments and publish research articles in international journals and produce a research project in various disciplines or real industrial and research scenarios.
4. Contribute to e-Science related industries or social studies.
5. Demonstrate the ability to publish articles.

Associated Assessment Criteria for Exit Level Outcome 1

• Apply various computational and statistical methods to analyses scientific and business data.
• Assess the suitability of methods for data collection and use.
• Implement state-of-the-art machine learning methods efficiently using high-performance computing platforms.
• Undertake creative work, making systematic use of various investigation techniques to discover new knowledge.
• Report results in a clear and understandable manner.
• Analyse scientific and industrial data to originate new applications and support decision making.

Associated Assessment Criteria for Exit Level Outcome 2

• Focus on the core areas of data science.
• Demonstrate inclination towards further studies (e.g. PhD studies in Computer Science or Statistics to deepen one's knowledge of the field and research methods).
• Teach e-Science courses at undergraduate and postgraduate levels.

Associated Assessment Criteria for Exit Level Outcome 3

• Demonstrate the understanding of the internationalised curriculum.
• Participate in international research.
• Provide evidence of publications in international journals.

Associated Assessment Criteria for Exit Level Outcome 4

• Choose an MSc project that is case-based showing the practical application of the skills in various disciplines or real industrial and research scenarios.
• Participate in interdisciplinary research.
• Attend workshops and conferences in e-Science.

Associated Assessment Criteria for Exit Level Outcome 5

• Publish articles.
• Supervise postgraduate research projects.

Integrated Assessment

Formative and summative assessment is as per the above-stated tools and formulae.

The qualification includes a compulsory Master's dissertation. The dissertation will focus on a data science problem and on applying the knowledge one will have learned during the course of doing modules in solving the selected issue. The problem one identifies can be theoretical or practical, but the dissertation must always have a research component and a scientific basis.

In the dissertation, one must demonstrate his/her ability to think scientifically, choose appropriate research methods and produce a well written scientific project. The dissertation must contain a definition of the research questions, a review of the relevant literature, statement of the problem, and methods adopted for the purpose constructing a theoretical, constructive or empirical basis for the development of solutions to stated research questions. A supervisor will be appointed to oversee one's dissertation. Learners must have regular meetings with his/her supervisor to ensure that his/her work is progressing smoothly and on schedule. The dissertation is worth 90 Credits, roughly corresponding to one year of full-time studies.