Page 177 - FINAL_HANDBOOK_20242025

Faculty of Science Handbook, Academic Session 2024/2025

unsupervised learning. SIT3017
Application of linear STATISTICAL LEARNING
discriminant analysis, AND DATA MINING
classification and regression
trees for supervised learning. This course prepares students
for applied work in data science
Assessment: by building on students’
Continuous Assessment: 40% foundations of data science
Final Examination: 60% skills. Students will learn
advanced methods in statistical
learning and data mining, using
SIT3016 appropriate computing tools
GENERALIZED LINEAR such as R. The strengths of the
MODELS diversity of approaches are
illustrated through analyses of
Introduction to generalized real world data sets covering
linear model based on the commonly encountered data
exponential family. For types.
example, multiple linear
regression for normal data, Exploratory analyses:
logistic regression for binary dimensional reduction methods
data, Poisson regression for such as principal components
counts, log linear for analysis and linear discriminant
contingency table, and gamma analysis. Feature selection.
regression for continuous non- Supervised learning: artificial
normal data. neural networks, k-nearest
Study the theory of GLM neighbours, logistic regression,
including estimation and naïve-Bayes, classification and
inference. regression trees, or support
vector machine. Ensemble
Introduction to fitting GLM in R. methods: bagging, random
forest, and boosting.
Focus on the analysis of data: Unsupervised learning: K-
binary, count and continuous, means and hierarchical
model selection, model clustering.
evaluation, interpretation,
prediction and residual Assessment:
analysis. Continuous Assessment: 50%
Final Examination: 50%
Assessment:
Continuous Assessment: 40%
Final Examination: 60%

177