Department of Decision Sciences

Prof MD Jankowitz

College of Economic and Management Sciences
School of Economic and Financial Sciences
Department: Decision Sciences
Associate Professor
Tel: 012 433 4731
E-mail: jankomd@unisa.ac.za

Qualifications

  • BSc(UOFS)
  • BSc(Hons)(US)
  • MSc(US)
  • HED(Unisa)
  • PhD(US)

Currently teaching

  • Forecasting

Fields of academic interests

  • Time Series and the Smoothing of Time Series
  • Nonlinear smoothing methods, especially Lower-Upper-Lower-Upper (LULU) smoothers compared to other nonlinear smoothers such as the median smoother and variations of it

Field of Specialisation

  • Statistics

Journal articles

Refereed/peer-reviewed articles in journals

  • Conradie, WJ, de Wet, T and Jankowitz, MD (2009) Performance of nonlinear smoothers in signal recovery.
    Applied Stochastic Models in Business and Industry, 25, 425-444
  • Conradie, WJ, de Wet, T and Jankowitz, M (2006) Exact and asymptotic distributions of LULU smoothers.
    Journal of Computational and Applied Mathematics, 186(2006), 253-267
  • Conradie, WJ, de Wet, T and Jankowitz, MD (2005) An overview of LULU smoothers with application to financial data.
    Journal for Studies in Economics and Econometrics (S.E.E.), Vol 29, No 1, 97-121

Paper presentations

Conference outputs:

  • ORSSA conference 2013. Nonlinear smoothing: A comparison of median, hybrid and LULU smoothers.
  • SASA conference 2010. Real-Time signal extraction using nonlinear smoothing.
  • International Symposium on Business and Industrial Statistics 2010 (ISBIS2010), Portoroz, Slovenia. Monitoring structural change in time series models.
  • Stellenbosch Statistical Symposium 2009. Nonlinear smoothing of special functions with non-Gaussian noise.
  • SASA conference 2008. The preservation of edges in nonlinear smoothing.
  • International Symposium on Business and Industrial Statistics 2008 (ISBIS2008), Prague, Czech Republic. Smoothing spatially variable functions using nonlinear smoothers.
  • International Federation of Operational Research Societies (IFORS) 2008 conference. Sandton, South Africa. New developments in nonlinear smoothing.
  • SASA conference 2007. Practical applications of LULU smoothers.
  • International Symposium on Business and Industrial Statistics 2007 (ISBIS2007), University of Azores, Portugal. Performance of Lower-Upper-Lower-Upper nonlinear smoothers in signal recovery.
  • SASA conference 2006. Smoothing special functions - Blocks, Bumps, HeaviSine and Doppler.
  • SASA conference 2005. Signal recovery with LULU smoothers.
  • Fourth International Symposium on Business and Industrial Statistics (ISBIS4), Cairns, Australia 2005. Nonlinear (LULU) Smoothers for Time-Series Data.
  • SASA conference 2004. The class of order based nonlinear smoothers.
  • SASA conference 2001. Income Protection Quality Control Sampling Plan.
  • SASA conference 1991 A comparison of hierarchical and non-hierarchical clustering techniques by an empirical study.

Projects

Community engagement:

  • Terminology Hub project of the Dept. of Decision Sciences