6CM521 — Frontiers in AI and Data Science

Overview

This module surveys the frontiers of artificial intelligence and data science, combining hands-on Python programming with an exploration of modern machine learning and deep learning techniques. Topics progress from classical ML through deep neural networks, explainable AI, reinforcement learning, and large generative models, with practical lab sessions throughout.

Topics

• Python for data science: NumPy, pandas, and visualisation
• Supervised learning: regression, classification, and decision trees
• Unsupervised learning: clustering and dimensionality reduction
• Artificial neural networks and deep learning
• Explainable AI (XAI)
• Reinforcement learning
• Generative AI and large language models
• Fine-tuning with LoRA

Companion notebooks

Three open courses on the teaching page cover most of this module in hands-on notebooks that run in Google Colab. They are general courses, not tied to this module, but the table shows where to look for each topic.

Topic	Open course	Notebooks
Python for data science	From Python to Deep Learning and Fine-Tuning	1 to 7
Supervised learning	From Python to Deep Learning and Fine-Tuning	8, 9, 10
Unsupervised learning	From Python to Deep Learning and Fine-Tuning	11
Neural networks and deep learning	From Python to Deep Learning and Fine-Tuning	12 to 15
Reinforcement learning	From Random Moves to Winning Agents	1 to 7 for the foundations, 8 to 15 for deep reinforcement learning
Generative AI and large language models	From Python to Deep Learning and Fine-Tuning	16, 17
	From Words to Meaning	12, 15, 18
Fine-tuning	From Python to Deep Learning and Fine-Tuning	18

Notebook 18 covers fine-tuning in general, which LoRA builds on. Explainable AI and LoRA do not have a notebook of their own yet.