¹û¶³Ó°Ôº

Aims:

Responsible AI is a combination of principles, practices and tools that enable the deployment of AI technologies in an ethical, transparent, secure, and accountable manner. This module covers the implications of Artificial Intelligence and introduces novel research strategies for building accountable, transparent, and responsible intelligent machines. Among others, this course introduces concepts related to risk and decision making with AI, fair and unbiased machine learning algorithms, safety and trust in human-machine systems, policymaking with and for AI and transparency and interpretability of AI technology, all current open challenges for the artificial intelligence community and with a crucial role to play in building a sustainable society.

The aims of the module are to:

• Support students in the development of a breadth of knowledge and understanding of the implications of AI technologies.
• Capacitate ML and AI practitioners in the development and deployment of robust and trustworthy AI systems.
• Provide an applied context for the use of fundamental concepts and latest research trends in responsible AI.

Intended learning outcomes:

On successful completion of the module, a student will be able to:

1. Understand and have assimilated the fundamental principles, theory, and approaches for building transparent, responsible, human-centred, and accountable intelligent systems.
2. Understand and discuss the broad impact of artificial intelligence technologies on the real-world.
3. Evaluate the quality (in terms of fairness, algorithmic bias, and robustness) of different machine learning models.
4. Develop and validate main algorithmic practices to build responsible AI systems.

Indicative content:

The following are indicative of the topics the module will typically cover:

• Practices for designing human-centred AI systems.
• Algorithmic fairness.
• Interpretable Machine Learning.
• Privacy-preserving algorithms.
• Security, ethics & policy.

Requisites:

To be eligible to select this module as optional or elective, a student must: (1) be registered on a programme and year of study for which it is a formally available; (2) have some machine learning background, for example from Supervised Learning (COMP0078), Introduction to Machine Learning (COMP0088), Foundations of Artificial Intelligence (COMP0186), or Deep Representations and Learning (COMP0188); and (3) have some programming skills (preferably Python).