There is an explosion in data generation and data collection due to improvements in sensing technologies and business processes. Extracting meaningful knowledge from large amounts of data has become a priority for businesses as well as scientific domains. Machine learning provides core underlying theory and techniques to data analytics, where algorithms iteratively learn from data to uncover hidden insights. In this unit, students will develop in-depth understanding of machine learning techniques that are applicable to both scientific and business data. The topics covered by the unit include supervised classification, unsupervised classification, regression, support vector machines, decision trees, random forests, dimensionality reduction, artificial neural networks, deep neural networks, autoencoders, and reinforcement learning.

Students are able to (1) explain the role of machine learning in knowledge extraction; (2) explain the difference between supervised and unsupervised learning algorithms; (3) demonstrate a systematic knowledge of algorithmic machine learning approaches; (4) produce practical implementations of machine learning solution for a real-world dataset; (5) analyse data datasets from the perspective of machine learning; and (6) evaluate what deep learning is, what makes it work or fail, and critique where it should be applied.