2019/12/09 by Jonas I. Liechti, Liechti, Jonas I., Sebastian Bonhoeffer +1 · 1 citation
Computer Science · Economics, Econometrics and Finance · Physics and Astronomy · #62-07 #Complex Network Analysis Techniques #Complex Systems and Time Series Analysis #Data Visualization and Analytics #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML)
paper · pdf · doi:10.48550/arxiv.1912.04261
openalex publication_date 2019/12/09 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
The last decades have not only been characterized by an explosive growth of\ndata, but also an increasing appreciation of data as a valuable resource. Their\nvalue comes with the ability to extract meaningful patterns that are of\neconomic, societal or scientific relevance. A particular challenge is the\nidentification of patterns across time, including those that might only become\napparent when the temporal dimension is taken into account. Here, we present a\nnovel method that aims to achieve this by detecting dynamic clusters, i.e.\nstructural elements that can be present over prolonged durations. It is based\non an adaptive identification of majority overlaps between groups at different\ntime points and accommodates the transient decompositions in otherwise\npersistent dynamic clusters. Our method enables the detection of persistent\nstructural elements with internal dynamics and can be applied to any\nclassifiable data, ranging from social contact networks to arbitrary sets of\ntime stamped feature vectors. It represents a unique tool to study systems with\nnon-trivial temporal dynamics and has a broad applicability to scientific,\nsocietal and economic data.\n