Emerging Topics in Integrated Machine Learning Systems

Teaching Overview

I delivered a guest lecture for the MSc course Emerging Topics in Integrated Machine Learning Systems within the Integrated Machine Learning Systems MSc at UCL’s Department of Electronic and Electrical Engineering. My lecture, titled “Quantum Artificial Intelligence: From Optimisation and Machine Learning to Solving Real-World Problems,” introduced students to the emerging field of Quantum Artificial Intelligence (QAI) and its relevance to next-generation machine learning systems.

The lecture was designed for MSc students with backgrounds in machine learning, electronic and electrical engineering, data systems, and applied computing. It provided an accessible but technically grounded overview of how quantum computing may contribute to optimisation, machine learning, sampling, and decision-making workflows in realistic scientific and industrial settings.

The session connected Quantum Artificial Intelligence to the broader theme of integrated machine learning systems, including hardware/software co-design, hybrid computational workflows, scalable AI infrastructure, and emerging computing paradigms.

Lecture Presented

Quantum Artificial Intelligence: From Optimisation and Machine Learning to Solving Real-World Problems

Lecture Abstract

Quantum Artificial Intelligence (QAI) combines quantum computing and artificial intelligence to explore new approaches for solving complex optimisation and machine learning problems. This lecture presented an accessible and practical overview of QAI, beginning with quantum optimisation and extending to quantum machine learning and hybrid classical–quantum approaches.

The talk highlighted promising real-world applications, current technical limitations, and the conditions under which QAI may deliver value in the near term. Application areas included logistics, finance, healthcare, energy systems, wireless networks, and scientific data analysis. The session emphasised a realistic view of QAI, focusing not only on potential quantum advantages but also on benchmarking, hardware constraints, workflow design, and integration with classical machine learning systems.

Teaching Contributions

As the guest lecturer, I designed and delivered a structured lecture covering the following topics:

1. Motivation: Why Quantum Artificial Intelligence?

Introduced the motivation for Quantum Artificial Intelligence and explained why optimisation, sampling, high-dimensional learning, and complex decision-making problems are central targets for quantum-enhanced workflows. The lecture framed QAI as an emerging computational paradigm rather than a direct replacement for classical AI.

2. Quantum Optimisation

Presented the core ideas behind quantum optimisation, including QUBO formulations, Ising models, quantum annealing, variational quantum optimisation, and hybrid quantum–classical optimisation loops. Examples were discussed in the context of logistics, scheduling, finance, energy systems, and industrial decision problems.

3. Quantum Machine Learning

Introduced major families of quantum machine learning methods, including quantum kernels, variational quantum circuits, quantum neural networks, and quantum-enhanced feature maps. The lecture discussed how quantum models differ from conventional machine learning pipelines and where they may be relevant.

4. Hybrid Classical–Quantum Workflows

Explained how near-term quantum algorithms are typically embedded within classical workflows. Topics included data encoding, circuit execution, measurement, classical post-processing, optimisation loops, and the practical impact of quantum hardware noise and sampling limitations.

5. Real-World Applications

Discussed realistic application domains where QAI is being explored, including:

• logistics and routing optimisation;
• portfolio optimisation and financial modelling;
• healthcare and high-dimensional biomedical data;
• energy systems and power-grid optimisation;
• wireless networks and resource allocation;
• scientific machine learning and simulation workflows.

6. Benchmarking and Practical Limitations

Emphasised the importance of rigorous benchmarking, fair comparison with classical baselines, hardware-aware evaluation, and time-to-solution or time-to-quality metrics. The lecture also covered current limitations, including qubit count, noise, data-loading overhead, limited circuit depth, and the difficulty of demonstrating practical quantum advantage.

7. From Research to Impact

Concluded with a discussion of how QAI can move from academic research to real-world deployment. The lecture highlighted the need for interdisciplinary collaboration across quantum algorithms, machine learning, high-performance computing, domain expertise, and systems engineering.

Learning Outcomes

By the end of the lecture, students were expected to be able to:

• explain the basic motivation behind Quantum Artificial Intelligence;
• distinguish between quantum optimisation, quantum machine learning, and hybrid quantum–classical workflows;
• identify realistic application areas for QAI;
• understand the main limitations of near-term quantum computing;
• evaluate QAI claims critically using benchmarking and systems-level reasoning;
• relate QAI to broader integrated machine learning systems and emerging AI infrastructure.

Teaching Approach

The lecture combined conceptual explanation, practical examples, and critical discussion. Rather than presenting QAI as a purely theoretical topic, the session focused on how quantum methods interface with modern machine learning systems and what is required for credible deployment.

The teaching style was designed to be accessible to a broad MSc audience while still introducing technically meaningful concepts such as variational circuits, QUBO modelling, quantum kernels, hybrid optimisation loops, and benchmarking methodology.

Relevance to Integrated Machine Learning Systems

This teaching engagement connected Quantum Artificial Intelligence to the broader engineering of integrated machine learning systems. In particular, it linked quantum methods with:

• advanced computational infrastructure;
• data-driven modelling workflows;
• hardware/software co-design;
• scalable AI systems;
• emerging computing paradigms;
• responsible evaluation of new machine learning technologies.

The lecture demonstrated my commitment to teaching frontier topics in a way that is technically rigorous, industry-aware, and accessible to students preparing for careers in advanced AI systems, engineering, research, and technology translation.