Course Descriptions (AY2026/27 Intake)

Deep learning is one of the bleeding-edge technologies of machine learning. It is a neural network used to establish and simulate the human brain for analytical learning and to interpret data by imitating the mechanism of the human brain. Deep learning is widely used in computer vision, speech recognition, natural language processing, and other fields. This course aims at providing an intensive understanding and hands-on experience of the existing deep learning approaches. The topics will cover how to select deep neural networks, how to design deep neural networks, and how to train and optimize the neural networks for practical applications. The course will cover deep neural network models, including convolutional neural networks, recurrent neural networks, long short-term memory networks, deep residual networks, generative adversarial networks, attention-based models, adversarial learning models, and training techniques including dropout, batch normalisation, selection of activation functions and so on. TensorFlow, Pytorch, or other state-of-the-art deep learning tools will be introduced and applied to solve different classes of problems with huge datasets in business domains.

This course is about AI agent and large language models (LLMs) infrastructure. Agentic AI focuses on the design, theory, and construction of autonomous, goal-driven AI agents and multi-agent systems powered by LLMs. This course will discuss fundamental concepts such as foundations of LLMs, agent architecture, agent planning and tooling, agent memory and context, reasoning loops, collaboration in multi-agent environments, and agent evaluation and benchmarking. Meanwhile, we will discuss limitations and potential risks of current LLM agents, and share insights into directions for further improvement. The course emphasizes both theoretical foundations and hands-on practical implementations, including building agents using open-sourced modern agentic frameworks.

This course introduces the foundations of data science and programming, covering language syntax, program control, functions, and file handling. It explores AI-Assisted Python programming concepts such as classes, inheritance, and polymorphism, alongside the data science process and its applications. Students will gain practical experience with NumPy, SciPy, and Pandas for computational analysis and develop troubleshooting skills using generative AI toolThis course introduces the foundations of data science and programming, covering language syntax, program control, functions, and file handling. It explores AI-Assisted Python programming concepts such as classes, inheritance, and polymorphism, alongside the data science process and its applications. Students will gain practical experience with NumPy, SciPy, and Pandas for computational analysis and develop troubleshooting skills using generative AI tools.This course introduces the foundations of data science and programming, covering language syntax, program control, functions, and file handling. It explores AI-Assisted Python programming concepts such as classes, inheritance, and polymorphism, alongside the data science process and its applications. Students will gain practical experience with NumPy, SciPy, and Pandas for computational analysis and develop troubleshooting skills using generative AI

This course provides an engaging and accessible introduction to the essential tools and design principles behind cutting-edge AI and agentic systems, specifically tailored for non-technical learners. Through dynamic lectures and hands-on labs, students will explore how AI models are built, deployed, and integrated into real-world applications. The course emphasizes practical insights into Machine Learning Operations (MLOps), model lifecycle management, tracking, and service scaling. Students will also learn about the development and integration of agentic AI into business and decision-making contexts. By the end of the course, students will be equipped with the conceptual and managerial skills to bridge technical and business domains, enabling them to design, evaluate, and oversee AI-driven projects and agent workflows in diverse industries.

This course provides a focused and in-depth study of core machine learning methods that underpin modern artificial intelligence (AI) systems. Rather than surveying all stages of industrial machine learning in equal breadth, the course concentrates on the central modeling and evaluation components of the machine learning lifecycle, with selected exposure to data preparation, deployment considerations, and ethical issues where they directly affect model design and performance.

Students will develop a deep understanding of supervised and unsupervised learning, statistical evaluation methodologies, dimensionality reduction, Bayesian approaches, and foundational neural networks. These topics are examined from a system-level perspective, emphasizing how modeling choices, data characteristics, and evaluation strategies interact within an AI system. Through theory-driven instruction and carefully scoped hands-on projects, students learn to design, analyze, and reason about machine learning components as part of larger AI systems, preparing them for advanced study or professional practice.

This course provides a rigorous and interdisciplinary examination of the ethical, legal, and societal implications of artificial intelligence. It examines how agentic AI systems influence decision-making and execute autonomous actions across domains such as governance, communication, security, commerce, and personal life, and analyzes the ethical challenges arising from their design, deployment, and use. The course introduces students to major ethical theories and global AI governance frameworks, and applies them to case studies involving algorithmic bias, data privacy, surveillance, misinformation, autonomous systems, agentic AI alignment, and emerging forms of intelligent automation.

The course emphasizes ethics-by-design, responsible innovation, and risk assessment approaches that support the development of trustworthy, well-aligned, and human-centred AI systems. Students are expected to develop the ability to critically evaluate AI technologies and policies, articulate well-reasoned ethical judgments regarding machine agency and human oversight, and engage responsibly with the societal impacts of artificial intelligence.

This course introduces the principles, methods, and challenges involved in designing, analyzing, and deploying trustworthy artificial intelligence systems, including autonomous agentic AI. Building foundations in machine learning, probability, and optimization, the course examines privacy, robustness, fairness, transparency, security, and accountability in modern AI models and agent-based frameworks. Students will study real-world failures and attacks on AI systems, explore technical defenses such as differential privacy and adversarial training, and critically evaluate the ethical and societal implications of deploying goal-driven AI agents. The course emphasizes conceptual understanding, applied reasoning, and research-oriented analysis of trustworthy AI technologies and autonomous workflows.

This course provides students with the opportunity to demonstrate innovative abilities and initiatives in Agentic AI Systems projects. Students will work collaboratively in groups to complete a substantial practical project under the guidance and supervision of a faculty member. The course emphasizes the integration and application of data science and artificial intelligence techniques, including agentic AI, to address complex, real-world scenarios. It also serves as a platform for students to present and share novel investigations of academic and/or industrial problems using data science and AI knowledge. Projects may involve collaborations with industry partners, applied research, or the advanced development of AI-enabled and autonomous agentic systems. Supervisors will be assigned through a coordinated allocation process managed by the Programme Director, taking into account: student project preferences and thematic alignment with faculty expertise, supervisor’s workload balance and any declared or potential conflicts of interest. Where conflicts arise (e.g., prior employment relationships, family connections, or direct industry sponsorship involving a supervisor), alternative supervision or co-supervision arrangements will be implemented. Assessment moderation mechanisms, including second marking or panel review of presentations, will also be used to ensure fairness and academic integrity.

This course is designed to describe the advanced concepts and principles of data management for business. Various types of databases will be discussed in this course, such as object-oriented, relational, document-oriented, NoSQL, and New SQL. Popular database management systems such as Microsoft SQL Server and/or Oracle will be described. Topics include data models (ER, relational, and others); query language (Structure Queries Language); management of semi- structured and complex data; NoSQL databases. It also covers the essential concepts, options, and best practices for data administration, data protection, privacy control, user security and management, and system configurations. It addresses topics about the general concepts of data disaster recovery, planning, and procedures.

In an optimization problem, one seeks to minimise or maximise an objective function or a number of objective functions with real, integer, and/or discrete variables, subject to constraints on the variables. Optimization refers to the study of these problems, their properties, the development and implementation of algorithms to solve these problems, and the application of these algorithms to real-world problems. In this course, advanced artificial intelligence algorithms such as multi/many objectives optimization algorithms, genetic algorithms, evolution strategies, ant colony optimization, particle swarm algorithms, differential evolution, and other meta-heuristic methods will be discussed. These methods are able to find optimal or near-optimal solutions for challenging optimization problems. This course will also describe some real-world applications that use these algorithms to handle difficult real-world problems.

Blockchain, as a decentralised open ledger, has proven to be a phenomenal success. This ground-breaking technique holds a huge promise in various fields, digital identification, data marketing, cryptocurrencies like bitcoin, etc. This course introduces students the fundamentals of blockchain, distributed ledger technology, alternative consensus, smart contracts and security, and cryptocurrencies. Case studies of cryptocurrencies and examples of application (e.g., Bitcoin) will be also elaborated. Students will understand the impact of blockchain technologies on financial services and other industries through assignments and projects.

With the rapid development of cloud resources, majority of traditional data centers has been replaced by cloud platforms (such as Amazon Web Services, Microsoft Azure, Google Cloud Platform) due to the limited bandwidth of resource. A qualified data scientist needs to equip with cloud computing skills, learning to perform a series of tasks in the data pipeline on the cloud such as data acquisition, data cleansing, data transformation and data mining, as well as model training and testing. This course aims to examine the latest trends of cloud computing and provide students with the fundamental knowledge of cloud computing. Students will learn the best practices in deploying and implementing cloud computing applicable for unique business requirements. The course discusses the conceptual topics of cloud technologies and provide hands-on experience through projects utilizing public cloud infrastructures. Topics include cloud delivery models (SaaS, PaaS, and IaaS); Cloud computing overview; Public cloud infrastructure, On-demand self-service, and resource pooling; rapid elasticity; measured service; cloud storage architecture (data distribution, durability, consistency, and redundancy); data deduplication; cloud security issues; case studies of current cloud computing platforms.

This course is an introduction to Natural Language Processing (NLP). It covers a brief overview of the field, including the cutting-edge text processing tasks (e.g., text summarization, named entity recognition, document classification, etc.), their computational problem setting and general thoughts of methodologies. State-of-the-art techniques will also be discussed, including generative sequence-to-sequence models, multimodal data modelling (e.g., image-to-text, video/audio-to-text), chatbot, question-answering system, topic modelling, etc.

This course will introduce the techniques for visual data processing and analysis. Topics include image processing and analysis in spatial and frequency domains, image restoration and compression, image segmentation and registration, morphological image processing, representation and description, feature description, face recognition, iris recognition, fingerprint recognition, image analysis topics, such as medical image analysis.

This course explores the principles and practices of AI user experience (UX) design and data visualization, focusing on how to create interfaces and visualizations that are effective, transparent, and intuitive for diverse users. UX for AI systems emphasizes the unique challenges of designing interactions with AI-driven functionality and autonomous agentic AI, where trust, interpretability, user feedback, inclusivity, shared agency, and accountability are central. Data visualization, by contrast, focuses on how AI outputs, agent decision-making processes, and complex datasets are represented visually, highlighting clarity, accuracy, storytelling, cognitive load management, and ethical communication. While both share common goals of interpretability and user-centeredness, UX addresses how people use, respond to, and collaborate with agentic AI systems, whereas data visualization addresses how AI results and data are presented and understood. By integrating these complementary principles, students will gain the skills to design human-centered AI interfaces and agent-driven workflows that foster trust and accessibility, alongside impactful data communication strategies that make AI insights actionable and meaningful.

This course provides a comprehensive exploration of reinforcement learning (RL), a core area of artificial intelligence where agents learn to make sequential decisions by interacting with dynamic environments. Students will study key concepts including Markov decision processes (MDPs), dynamic programming, temporal-difference learning, policy gradients, and deep reinforcement learning. Advanced topics such as model-based RL, multi-agent systems, and applications to robotics will also be covered. Emphasis will be placed on practical applications in robotics, computer vision, and large language model (LLM) training. Through hands-on labs, students will implement RL algorithms and apply them to simulated and real-world problems, equipping them with the skills to design, analyze, and apply reinforcement learning techniques in diverse AI domains.