David P. Hofmeyr

David P. Hofmeyr

Auto-generated profile, based on publicly available information

David P. Hofmeyr

Department of Statistics and Actuarial Science, Stellenbosch University, Stellenbosch, South Africa

Research Focus & Metrics

Academic Career Length

12Years (2015-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
3 February 2023 in Geoderma
Verified
Journal Article
Open access
SWStephan van der Westhuizen
Stephan van der Westhuizen
GHGerard Heuvelink
Gerard Heuvelink
DHDavid P. Hofmeyr
David P. Hofmeyr

15 September 2016 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Verified
Journal Article
Closed access
DHDavid P. Hofmeyr
David P. Hofmeyr
Minimum normalised graph cuts are highly effective ways of partitioning unlabeled data, having been made popular by the success of spectral clustering. This work presents a novel method for learning hyperplane separators which minimise this graph cut objective, when data are embedded in Euclidean space. The optimisation problem associated with the proposed method can be formulated as a sequence of univariate subproblems, in which the optimal hyperplane orthogonal to a given vector is determined. These subproblems can be solved in log-linear time, by exploiting the trivial factorisation of the exponential function. Experimentation suggests that the empirical runtime of the overall algorithm is also log-linear in the number of data. Asymptotic properties of the minimum cut hyperplane, both for a finite sample, and for an increasing sample assumed to arise from an underlying probability distribution are discussed. In the finite sample case the minimum cut hyperplane converges to the maximum margin hyperplane as the scaling parameter is reduced to zero. Applying the proposed methodology, both for fixed scaling, and the large margin asymptotes, is shown to produce high quality clustering models in comparison with state-of-the-art clustering algorithms in experiments using a large collection of benchmark datasets.

1 March 2022 in Spatial Statistics
Verified
Journal Article
Open access
SWStephan van der Westhuizen
Stephan van der Westhuizen
GHGerard B.M. Heuvelink
Gerard B.M. Heuvelink
DHDavid P. Hofmeyr
David P. Hofmeyr
LPLaura Poggio
Laura Poggio

1 September 2020 in Computational Statistics & Data Analysis
Verified
Journal Article
Closed access
DHDavid P. Hofmeyr
David P. Hofmeyr

23 July 2019 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Verified
Journal Article
Closed access
DHDavid P. Hofmeyr
David P. Hofmeyr
This paper presents new methodology for computationally efficient evaluation of univariate kernel sums. It is shown that a rich class of kernels allows for exact evaluation of functions expressed as a sum of kernels using simple recursions. Given an ordered sample the computational complexity is linear in the sample size. Direct applications to the estimation of denisties and their derivatives shows that the proposed approach is competitive with the state-of-the-art. Extensions to multivariate problems including independent component analysis and spatial smoothing illustrate the versatility of univariate kernel estimators, and highlight the efficiency and accuracy of the proposed approach. Multiple applications in image processing, including image deconvolution; denoising; and reconstruction are considered, showing that the proposed approach offers very promising potential in these fields.

1 December 2015
Verified
Conference Paper
Closed access
DHDavid Hofmeyr
David Hofmeyr
NPNicos Pavlidis
Nicos Pavlidis
The notion of cluster ability is often used to determine how strong the cluster structure within a set of data is, as well as to assess the quality of a clustering model. In multivariate applications, however, the cluster ability of a data set can be obscured by irrelevant or noisy features. We study the problem of finding low dimensional projections which maximise the cluster ability of a data set. In particular, we seek low dimensional representations of the data which maximise the quality of a binary partition. We use this bi-partitioning recursively to generate high quality clustering models. We illustrate the improvement over standard dimension reduction and clustering techniques, and evaluate our method in experiments on real and simulated data sets.