To demonstrate in-depth understanding and knowledge of the concepts, theories and developments associated with the subject area, and critically and analytically discuss outcomes in a methodological, structured, logical and in-depth manner;

To demonstrate the ability to apply current tools and techniques of machine learning and data mining in suitable depth and at the appropriate level.

(LO1) Gain an in depth knowledge and principles in the areas of data mining and machine learning;

(LO2) Critically assess the strengths and weaknesses of various data mining and machine learning techniques from a practitioner/ user perspective;

(LO3) Be able to identify, formulate and solve problems arising from practical applications using data mining and machine learning principles and techniques.

(S1) Adaptability
Students will develop adaptability by engaging with case studies and assignments to understand the application of data mining and machine learning techniques.

(S2) Problem solving skills
Students will develop problem solving skills through practicing exercises and undertaking assignments.

(S3) Commercial awareness
Students will develop knowledge of commercial contexts of machine learning techniques and their practical issues.

(S4) Organisation skills
Students will develop time management skills by meeting deadlines of class discussion tasks and assignments.

(S5) Communication skills
Students will develop communication skills by engaging with case studies, report writing and working in groups.

(S6) IT skills
Students will develop IT skills and perform programming exercises during lab sessions.

(S7) International awareness
Students will develop international awareness through case studies of data mining in an international context.

(S8) Lifelong learning skills
Students will develop skills of lifelong learning through self-directed study of cases and reading materials, finding relevant information, and preparation for their assessments.

2 hour lecture x 12 weeks
1 hour seminar x 6 weeks
1 hour group learning x 6 weeks
114 hours self-directed learning

Concepts of data mining, machine learning, supervised learning, unsupervised learning, and Cross-Industry Standard Process Data Mining (CRISP-DM) model.

Data understanding and data preparation; data attributes and data types (nominal, ordinal, interval and ratio data); data quality issues (e.g. missing values, noise, outliers); data pre-processing including aggregation, sampling, feature selection, feature creation and discretisation; basic R programming language.

Data exploration; using R programming tools to explore dataset structure and summary statistics (e.g. median, mean, range and frequency); using R programming tools to visualise data (e.g. boxplot, scatter plot, 2D and 3D histograms) and identify outliers.

Classification modelling methods; induction process to build decision tree; deduction process to apply decision tree; the Hunt’s algorithm; impurity measures (Gini index and entropy); random forest classifier; rule-based classifier; practical issu es of classification (underfitting, overfitting, evaluation); using R programming tools to facilitate building decision trees and calculating impurity measures.

Clustering: types of clusters; types of clustering techniques; applications of clustering; K-means algorithm; methods for evaluating the clustering performance; limitations of K-means; hierarchical clustering algorithm; agglomerative clustering algorithm; approaches to define inter-cluster similarity; using R programming tools for clustering.

Mining association rules: introduction to association rule mining; types of association rules; a two-step approach to mine association rules; the use of Apriori algorithm; computational steps for rule generation; hash tree technique; ,recent applications of association rule mining technique; using R programming tools for association rule learning.

Fuzzy logic technique: the fuzziness of data; fuzzification and membership function determination; composition of fuzzy sets; fuzzy inference engine; defuzzification and rule firing; recent applications of fuzzy logic technique; other integrated-fuzzy approaches.