Aims

This module will provide students with practical knowledge and deeper insight into how cutting-edge data science techniques and methods, machine-learning and artificial-intelligence are being applied to address real-world problems faced by businesses and institutions. Students will also be introduced to the project realities faced by practitioners of data science in the workplace, and apply the professional skills required to tackle the delivery of real-world data science projects.

Outline Of Syllabus

Students will be challenged to undertake short projects, let by NICD data scientists, based on data science problems actively being explored in the workplace. The module leverages the project-based experience of the expert data scientists of the National Innovation Centre for Data (NICD). Course content will be derived from the practical experience of NICD data scientists working on a portfolio of collaborative projects with organisations of all sizes, to deliver solutions to a wide range of clients across private, public and third sector. Course content will be derived from real world challenges and the most appropriate data science technique(s) to apply will vary as there will be many possible avenues for students to explore. These may include:
-The use of Low code and No code solutions.
-Developing descriptive data products (e.g. reports, dashboards, etc).
-Applying statistical modelling techniques, such as linear or logistic regression.
-Machine learning approaches (e.g. Decision trees, Random Forests etc.).
-Large Language Models (LLMs) and National Language Processing (NLP).
-Deep learning on computer vision, image processing or audio processing.
-Application of generative-AI to create custom documents or audio-visual content.

Successful delivery of data science projects in the workplace is paramount. To understand this, course content will cover:
-Scoping and specifying innovative data science solutions, based on NICD experience.
-Student-led research into cutting-edge DS techniques, such as generative AI.
-Selection and application of identified techniques to real-world challenges.
-Hands-on experiential learning alongside NICD data scientists.
-Conveying findings and results successfully to both technical and non-technical audiences.
-Using data science in an ethical and responsible manner.
-The importance of self-reflection and self-guided learning to achieve personal growth and development.

The module will employ a 'flip learning' style, where after an initial introduction, working in small groups, students will be expected to direct their own investigation toward a solution to a particular case study, with guidance and supervision from NICD data scientists. The module will consist of three project blocks which are three-weeks in duration. Each project block will be based around a particular case-study or innovative data science technique, along with detailing how NICD approaches the task of upskilling businesses in the effective and successful use of data science.

Teaching Rationale And Relationship

The module will primarily employ a student-driven 'flip' learning style involving selected project case studies over three-week block periods. In an introductory session, students will be introduced to a particular business problem that NICD data scientists have faced, along with initial guidance as to how that problem may be addressed. Students will then work to:
-Examine and develop a sound understanding of the case study problem.
-Investigate and analyse data provided to establish an evidence base for decisions.
-Identify and research appropriate data science techniques, giving advantages and disadvantages of these.
-Select an appropriate technique from those examined, giving justification for doing so.
-Reflect on how work is progressing within their group and what their individual contribution has been.

The above should be accomplished before the week two session, wherein NICD staff will provide feedback and guidance on progress.

Students will then proceed to:
-Outline the technical steps, data processing and platform requirements for successful application of their chosen methodology.
-Specify what expected project outcomes might be, and what benefits and pitfalls may arise.
-Students are not expected to provide complete, functional, 'production ready' solutions in the three-week period, but credit will be given if students demonstrate awareness of how such could be achieved (e.g. by identifying appropriate testing procedures and deployment aspects).

Finally, students will convey the above information to NICD data scientists through presentation at the week 3 session, where further feedback and critique from NICD data scientists will be provided as formative assessment. NICD data scientists will also detail how the Innovation Centre tackled the problem, with discussion of the approach across groups.

There will be three case studies presented over the course of the module. An additional, interim lecture will also be provided by NICD data scientists, outlining the general approach the Innovation Centre takes to try to ensure client success with data science, and to introduce other project topics not covered by the three main case studies (e.g. newer AI techniques applied to active projects). The last week of the module will be allocated for summative assessment by formal group presentation on one of the topics covered in th ecase study blocks.

Assessment Rationale And Relationship

The bulk of the assessment mark is provided for the group presentation, which assesses the technical solution provided by the group, how practical and feasible that solution might be and how the solution fits the needs and goals of the client. The presentation also provides an indication of how cohesively the group has worked towards the solution and how clearly they are able to communicate their approach to the client (who may not necessarily by technically oriented). The short report gives the student a chance to reflect on what they have learnt and to provide their own comment on how they have found group working, how effective their collaboration with their peers has been and personally critique their own solution from the main presentation. The two assessments are intended therefore to be somewhat diverse (expressive group work vs. personal, internal reflection) and cover the multitude of intended learning and skill outcomes for the module.