01
Unbiased mixed-variable distance
Journal of Computational and Graphical Statistics, 1–12
(01) data analysis / methods / software
Distances,
structure,
and everything in between.
I work on multivariate data analysis, with particular interests in clustering, dimensionality reduction and mixed-type data.
Angelos Markos Professor of Data Analysis in Social Sciences Democritus University of Thrace · Greece
01 pairwise dissimilarities 02 low-dimensional structure 03 clustering 04 mixed data
02
Research space
Structure begins
with proximity.
My methodological work is centred on ways of defining, representing and exploiting relationships among observations and variables — especially when the data are heterogeneous.
conceptual mapproximity ≠ taxonomy
03
Selected research
Recent work
Selected publications on distance-based methods, clustering, dimensionality reduction, mixed-type data and measurement.
02
A deterministic information bottleneck method for clustering mixed-type data
Pattern Recognition 179(B), 113580
03
A general framework for implementing distances for categorical variables
Pattern Recognition 153, 115547
04
Dimensionality Assessment in Ordinal Data: A Comparison Between Parallel Analysis and Exploratory Graph Analysis
Frontiers in Psychology 15, 1359111
05
Benchmarking distance-based partitioning methods for mixed-type data
Advances in Data Analysis and Classification 17, 701–724
04
Statistical software
Methods
made usable.
Open-source implementations accompanying methodological work.
01
manydist
R package · CRANDistance-based learning for numerical, categorical, and mixed-type data.
↗
02
IBclust
R package · CRANInformation Bottleneck methods for clustering continuous, categorical, and mixed-type data.
↗
03
clustrd
R packageJoint dimension reduction and clustering in R.
↗
04
idm
R packageIncremental decomposition methods, including incremental PCA and MCA.
↗
05
Writing
Notes from
the archive.
Recovered posts from the previous amarkos.gr archive. Newer long-form writing lives at Celebrating Uncertainty.