Curriculum Information

Data analysis and preprocessing:

Exploratory data analysis; Handling missing data; Feature engineering techniques, including but not limited to: transformations, feature extraction, reduction and selection.

Supervised and Unsupervised learning:

Introduction to the core concepts of supervised and unsupervised learning while looking at classical and recent algorithms and their applications on real datasets.

Supervised learning: parametric, non-parametric, probabilistic and kernel-based classification and regression algorithms. Optimising the supervised learning models such as: linear, logistic and locally weighted regression. Overfitting, underfitting, parameter tuning, model selection and generalisation using widely used algorithms for supervised learning such as: Naïve Bayes, support vector machine and decision trees.

Supervised model evaluation methods and metrics.

Unsupervised learning: clustering techniques as unsupervised machine learning models and dimensionality reduction algorithms. Distance-based, hierarchical and density-based unsupervised learning models including hard and soft clustering models. Cluster evaluation metrics such silhouette coefficient. Unsupervised dimensionality reduction using algorithms such as principal component analysis and t-SNE.

Neural networks:

Artificial Neural Networks (ANNs) and a 'black box' model of classification; linear separability; Multilayer feedforward networks; methods for determining weights; training networks and error functions; regularization; deep neural networks; characteristics of other types of ANN; Practical application using datasets.

Extension and applications:

Comparison with other approaches including k- nearest neighbour approach. Boosting.