I
Ioannis Psorakis
Researcher at University of Oxford
Publications - 14
Citations - 1408
Ioannis Psorakis is an academic researcher from University of Oxford. The author has contributed to research in topics: Bayesian probability & Population. The author has an hindex of 11, co-authored 14 publications receiving 1243 citations. Previous affiliations of Ioannis Psorakis include Edward Grey Institute of Field Ornithology.
Papers
More filters
Journal ArticleDOI
Overlapping community detection using Bayesian non-negative matrix factorization
TL;DR: This work presents a probabilistic approach to community detection that utilizes a Bayesian non-negative matrix factorization model to extract overlapping modules from a network.
Journal ArticleDOI
Reality mining of animal social systems.
Jens Krause,Stefan Krause,Robert Arlinghaus,Robert Arlinghaus,Ioannis Psorakis,Stephen J. Roberts,Christian Rutz +6 more
TL;DR: Important issues concerning the collection of data on the social dynamics of almost entire populations of individuals, and their processing and analysis, are reviewed to identify the most promising approaches in the emerging field of 'reality mining'.
Journal ArticleDOI
Inferring social network structure in ecological systems from spatio-temporal data streams.
TL;DR: It is shown that established pair bonds are maintained continuously, whereas new pair bonds form at variable times before breeding, but are characterized by a rapid development of network proximity.
Book ChapterDOI
Dynamic Bayesian Combination of Multiple Imperfect Classifiers
TL;DR: In this paper, the authors explore Bayesian classifier combination, using the computationally efficient framework of variational Bayesian inference, using real data from a large citizen science project, Galaxy Zoo Supernovae, and show that their method far outperforms other established approaches to imperfect decision combination.
Journal ArticleDOI
Multiclass Relevance Vector Machines: Sparsity and Accuracy
TL;DR: It is demonstrated that mRVMs can produce state-of-the-art results on multiclass discrimination problems and this is achieved by utilizing only a very small fraction of the available observation data.