Journal of Quantitative Analysis in Sports · 2009 · 19 citations · 10 references
ComputingPsychologyHuman PerformanceStatisticsTime Series AnalysisCluster DetectionSerial AnalysisStatistical AnalysisTime Series ClusteringTime Series ForecastingTrend AnalysisCareer Progression AnalyticsCareer Progression PathwaysCluster Analysis For GroupingClusteringProfessional Career ProgressionTemporal Clustering
Time series clustering is used to show that, relatively, the career progression of Australian legend Sir Donald Bradman's test career as a batsman was most similar to West Indian Brian Lara. Consequently, it is likely his peak performance would have occurred while the Second World War disrupted all international cricket.Data from the 20 international cricketers who played in at least 70 innings over more than 17 years and averaged more than 40 runs per dismissal (as at January 1, 2009) is used to create a number of global measures that indicate the ebb and flow of a career. As is shown in this paper, this clustering methodology, proposed by Wang et al. (2006), generates instinctive clustering results and can be applied on different length time series.Utilizing the framework created for clustering, Bradman's batting average is estimated to be 105 if his career had been uninterrupted.
10
Clustering of time series data—a survey
T. Warren Liao · Pattern Recognition · 2005 · 2.4K citations
Hierarchical structure in financial markets
Rosario N. Mantegna · The European Physical Journal B · 1999 · 2K citations · Full text
On Clustering fMRI Time Series
Cyril Goutte, Peter Toft, Egill Rostrup et al. · NeuroImage · 1999 · 527 citations