Note: These modules are being offered for year 2022. Prospective honours students should not assume that these will be offered in 2023. To enquire about the modules being offered in the 2023 academic year, please contact us at stats@mandela.ac.za.
STAT401: Multivariate Statistical Methods
Lecturer: Dr J Hugo
This module aims to integrate theoretical and practical components of multivariate statistical analysis.
Core content: Matrix theory, partition matrices and Kronecker products, solutions of linear equations, latent roots and latent vectors of matrices, multivariate normal distribution, matrix normal distribution, moments, quadratic forms, estimation of parameters, the Wishart distribution, assessing multivariate normality, outliers in multivariate samples, tests on one or two mean vectors, profile analysis, tests on covariance matrices, discriminant analysis (description of group separation), classification analysis (allocation of observations to groups), principal component analysis, multivariate multiple regression.
STAT420: Quantitative Data Analysis
Lecturer: Dr WJ Brettenny
This module covers the theory and application of quantitative statistical method used for data analytics. R is used as the software of tuition.
Core content: Cross validation methods, bias-variance trade-off, linear regression methods, knn regression, nonparametric regression, local regression, knn classification, naïve Bayes, discriminant analysis, logistic regression, support vector machines, classification trees, random forests, model accuracy assessments, ROC curves, precision-recall curves, k-means clustering, hierarchical clustering.
STAT450: Selective Topics in Actuarial Statistics
This module covers the theory and application of quantitative statistical method used for data analytics. R is used as the software of tuition.
Core content: Cross validation methods, bias-variance trade-off, linear regression methods, knn regression, nonparametric regression, local regression, knn classification, naïve Bayes, discriminant analysis, logistic regression, support vector machines, classification trees, random forests, model accuracy assessments, ROC curves, precision-recall curves, k-means clustering, hierarchical clustering.
STAT480: Capita Selecta A (2021)
Lecturer: Prof I Litvine
This module will prepare students to use various mathematical and computational financial models for statistical analysis of financial data.
Core Content: Critical and coherent understanding and application of the underlying principles of: risk-return relationships, concepts of risk free investment, adjusted performance; shortfall risk, and safety rule, Sharpe ratio, security market line, Models for risk diversification, systematic and non-systematic risks. Markovitz efficient frontier, Models for asset price fluctuations, volatility, Introduction to efficient market theory, Introduction to asset valuations: arbitrage-free valuation principles, Models for arbitrage-free valuation of bonds and shares.
Note: The honours students will be required to complete a research project in addition to the aforementioned modules. The credit and NQF levels are provided below.
Module name |
Module code |
Credit value |
NQF level |
Core modules: These two modules are compulsory for those registered in either honours programme |
Project |
STAT400 |
30 |
8 |
Multivariate Statistical Methods |
STAT401 |
24 |
8 |
Other Modules |
Quantitative Data Analysis |
STAT420 |
24 |
8 |
Selective Topics in Actuarial Statistics |
STAT450 |
24 |
8 |
Capita Selecta A |
STAT480 |
24 |
8 |
Total Credits |
|
120 |
|