Archetypal analysis for machine learning and data mining

Mørup, Morten; Hansen, Lars Kai

doi:10.1016/j.neucom.2011.06.033

Cited by 188 publications

(224 citation statements)

References 21 publications

Supporting

Mentioning

222

Contrasting

Unclassified

Order By: Relevance

“…In recent years, AA has been used in different domains such as multidocument summarization Kononenko, 2013, 2014), the evaluation of scientists (Seiler and Wohlrabe, 2013), developmental psychology (Ragozini et al, 2017), biology (D'Esposito et al, 2012), market research and benchmarking (Li et al, 2003;Porzio et al, 2008;Midgley and Venaik, 2013), industrial engineering , e-learning (Theodosiou et al, 2013), machine learning problems (Mørup and Hansen, 2012), image analysis (Bauckhage and Thurau, 2009) and astrophysics (Chan et al, 2003).…”

Section: Related Workmentioning

confidence: 99%

Archetypoid analysis for sports analytics

Vinué

Epifanio

2017

Data Min Knowl Disc

View full text Add to dashboard Cite

We intend to understand the growing amount of sports performance data by finding extreme data points, which makes human interpretation easier. In archetypoid analysis each datum is expressed as a mixture of actual observations (archetypoids). Therefore, it allows us to identify not only extreme athletes and teams, but also the composition of other athletes (or teams) according to the archetypoid athletes, and to establish a ranking. The utility of archetypoids in sports is illustrated with basketball and soccer data in three scenarios. Firstly, with multivariate data, where they are compared with other alternatives, showing their best results. Secondly, despite the fact that functional data are common in sports (time series or trajectories), functional data analysis has not been exploited until now, due to the sparseness of functions. In the second scenario, we extend archetypoid analysis for sparse functional data, furthermore showing the potential of functional data analysis in sports analytics. Finally, in the third scenario, features are not available, so we use proximities. We extend archetypoid analysis when asymmetric relations are present in data. This study provides information that will provide valuable knowledge about player/team/league performance so that we can analyze athlete's careers.

show abstract

Section: Related Workmentioning

confidence: 99%

Archetypoid analysis for sports analytics

Vinué

Epifanio

2017

Data Min Knowl Disc

View full text Add to dashboard Cite

show abstract

“…Unlike EOFs archetypes are not required to be nested (Cutler and Breiman 1994;Bauckhage and Thurau 2009). However, like k-means clustering (and unlike EOFs) AA is invariant to translation and scaling and to rotational ambiguity (Morup and Hansen 2012).…”

Section: B Geometrical Interpretation Of the Archetypesmentioning

confidence: 99%

“…Most methods used to solve the AA problem (e.g., Seth and Eugster 2016;Morup and Hansen 2012;Porzio et al 2008;Bauckhage and Thurau 2009;Steinschneider and Lall 2015;Chan et al 2003) are essentially based on the original alternating constrained least squares algorithm of Cutler and Breiman (1994). The algorithm alternates between finding the optimal matrix A 5 (a ij ) for fixed archetypes Z and finding the optimal archetypes Z for fixed A.…”

Section: Numerical Solution Of Archetypes a Manifold-based Algorithmmentioning

confidence: 99%

“…This flexibility comes with a price, namely the interpretation (Morup and Hansen 2012), owing to, for example, the existence of efficient algorithms such as the singular value decomposition (SVD) and their nested nature. The latter refers to the fact that EOFs can be ordered naturally, where the set of the leading N EOFs are a subset of the set of the leading M EOFs for M .…”

Section: Introductionmentioning

confidence: 99%

“…For more references and further details on this topic the reader is referred to the recent review of Hannachi et al (2017). Similar to what we mentioned above, clustering is not the main focus here but is used to help understand the AA method (Morup and Hansen 2012). In clustering, for example, the original observations cannot be obtained as a direct combination of the cluster centroids.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Archetypal Analysis: Mining Weather and Climate Extremes

Hannachi

Trendafilov

2017

J. Climate

View full text Add to dashboard Cite

Conventional analysis methods in weather and climate science (e.g., EOF analysis) exhibit a number of drawbacks including scaling and mixing. These methods focus mostly on the bulk of the probability distribution of the system in state space and overlook its tail. This paper explores a different method, the archetypal analysis (AA), which focuses precisely on the extremes. AA seeks to approximate the convex hull of the data in state space by finding ''corners'' that represent ''pure'' types or archetypes through computing mixture weight matrices. The method is quite new in climate science, although it has been around for about two decades in pattern recognition. It encompasses, in particular, the virtues of EOFs and clustering. The method is presented along with a new manifold-based optimization algorithm that optimizes for the weights simultaneously, unlike the conventional multistep algorithm based on the alternating constrained least squares. The paper discusses the numerical solution and then applies it to the monthly sea surface temperature (SST) from HadISST and to the Asian summer monsoon (ASM) using sea level pressure (SLP) from ERA-40 over the Asian monsoon region. The application to SST reveals, in particular, three archetypes, namely, El Niño, La Niña, and a third pattern representing the western boundary currents. The latter archetype shows a particular trend in the last few decades. The application to the ASM SLP anomalies yields archetypes that are consistent with the ASM regimes found in the literature. Merits and weaknesses of the method along with possible future development are also discussed.

show abstract

Classification of social anhedonia using temporal and spatial network features from a social cognition fMRI task

Krohne

Wang

Hinrich

et al. 2019

Human Brain Mapping

View full text Add to dashboard Cite

Previous studies have suggested that the degree of social anhedonia reflects the vulnerability for developing schizophrenia. However, only few studies have investigated how functional network changes are related to social anhedonia. The aim of this fMRI study was to classify subjects according to their degree of social anhedonia using supervised machine learning. More specifically, we extracted both spatial and temporal network features during a social cognition task from 70 subjects, and used support vector machines for classification. Since impairment in social cognition is well established in schizophrenia‐spectrum disorders, the subjects performed a comic strip task designed to specifically probe theory of mind (ToM) and empathy processing. Features representing both temporal (time series) and network dynamics were extracted using task activation maps, seed region analysis, independent component analysis (ICA), and a newly developed multi‐subject archetypal analysis (MSAA), which here aimed to further bridge aspects of both seed region analysis and decomposition by incorporating a spotlight approach.We found significant classification of subjects with elevated levels of social anhedonia when using the times series extracted using MSAA, indicating that temporal dynamics carry important information for classification of social anhedonia. Interestingly, we found that the same time series yielded the highest classification performance in a task classification of the ToM condition. Finally, the spatial network corresponding to that time series included both prefrontal and temporal‐parietal regions as well as insula activity, which previously have been related schizotypy and the development of schizophrenia.

show abstract

Archetypal analysis for machine learning and data mining

Cited by 188 publications

References 21 publications

Archetypoid analysis for sports analytics

Archetypoid analysis for sports analytics

Archetypal Analysis: Mining Weather and Climate Extremes

Classification of social anhedonia using temporal and spatial network features from a social cognition fMRI task

Contact Info

Product

Resources

About