led by Professor Georgios Leontidis and Matthew Beddows, University of Aberdeen. This session is run in partnership with SUSTAIN CDT which is a UKRI AI Centre for Doctoral Training in Sustainable Understandable agri-food Systems Transformed by Artificial INtelligence.
This course will introduce core machine learning concepts, including the practical use of large language models for coding, with an emphasis on hands-on practice using relatable examples and discussing ethical implications.
The session will cover key aspects of data generation—primarily using synthetic data—while addressing a subset of tasks such as mental health survey data, personality survey data, sentiment analysis, time-series analysis, fairness audits, and psychometrics, among others. Participants will explore how machine learning can be applied to real-life tasks.
The session will blend theory and practice, covering topics and concepts related to supervised and unsupervised learning, data processing, model evaluation, privacy concerns, and ethical implications. Hands-on activities will be included, and participants will actively use large language models such as DeepSeek, ChatGPT, and Claude to understand their applications in coding exercises and visualisations.
By the end of the session, participants will have gained a basic understanding of machine learning models, their applications and evaluation, and the implications and limitations of using them for data analysis and predictive modelling.
No prior knowledge is required to follow the session. However, it would be beneficial to explore Google Colab in advance, particularly by setting up a Gmail account to access it and reviewing some of Colab’s tutorials that use Python. That said, there is no expectation that participants will have any prior experience with Python.