P
Pádraig Cunningham
Researcher at University College Dublin
Publications - 330
Citations - 12512
Pádraig Cunningham is an academic researcher from University College Dublin. The author has contributed to research in topics: Case-based reasoning & Feature selection. The author has an hindex of 56, co-authored 327 publications receiving 11187 citations. Previous affiliations of Pádraig Cunningham include Kern Medical Center & University of California, San Francisco.
Papers
More filters
k-Nearest Neighbour Classifiers
TL;DR: An overview of techniques for Nearest Neighbour classification focusing onMechanisms for assessing similarity (distance), computational issues in identifying nearest neighbours and mechanisms for reducing the dimen-sion of the data is presented.
Proceedings ArticleDOI
Tracking the Evolution of Communities in Dynamic Social Networks
TL;DR: A model for tracking the progress of communities over time in a dynamic network, where each community is characterised by a series of significant evolutionary events is used to motivate a community-matching strategy for efficiently identifying and tracking dynamic communities.
Proceedings ArticleDOI
Practical solutions to the problem of diagonal dominance in kernel document clustering
Derek Greene,Pádraig Cunningham +1 more
TL;DR: A selection of strategies for addressing the implications of diagonal dominance for unsupervised kernel methods in the task of document clustering are proposed, and their effectiveness in producing more accurate and stable clusterings is evaluated.
Journal ArticleDOI
Diversity in search strategies for ensemble feature selection
TL;DR: It is shown that, in some cases, the ensemble feature selection process can be sensitive to the choice of the diversity measure, and that the question of the superiority of a particular measure depends on the context of the use of diversity and on the data being processed.
Journal ArticleDOI
An analysis of the coherence of descriptors in topic modeling
TL;DR: Two out of three coherence measures find NMF to regularly produce more coherent topics, with higher levels of generality and redundancy observed with the LDA topic descriptors, suggesting that this may be a more suitable topic modeling method when analyzing certain corpora, such as those associated with niche or non-mainstream domains.