I specialise in the in the field of applied statistics. Specific applications include predictive models, image processing, financial models and stochastic process simulation. I am also well versed in various methods of statistical inference. I’m passionate about data mining and machine learning which I usually implement in R, Python, SAS or SPSS. I have 7 years’ experience in the financial/banking industry as a risk analyst which helped me understand how to implement theory in practice.
UT Research Information System
MSc: Applied statistics (mini-dissertation) (2012 – 2015)
The research for my mini-dissertation (Polygon Detection in Images) involved a thorough literature review of the field of mathematical morphology. In particular, we investigated various image smoothing and segmentation techniques. A new smoothing algorithm was proposed and an article was submitted to the Pattern Recognition Letter journal for review (2016). We combined the various techniques investigated in the literature review with our new algorithm and suggested that the new algorithm was the same or in some cases better than the current alternative. The analytics for this research were all done in Python and the document was written up in LaTex using JabRef for formatting the bibliography. This research was funded by the National Research Foundation (NRF).
The complete document can be found here: http://repository.up.ac.za/handle/2263/50823.
Part of the research (Polygon Detection in Images) was presented at the Statistical Association of South Africa (SASA) conference in 2013. We also presented a poster (Algorithms for Polygon Detection in Images), showcasing a condensed version of the research, at a conference hosted by the Pattern Recognition of South Africa in 2013. The poster also won second prize at a poster evening, at the University of Pretoria in 2013, for being one of the best posters submitted.