Hierarchical Time-Series Clustering for Data Streams⋆
نویسندگان
چکیده
This paper presents a time-series whole clustering system that incrementally constructs a hierarchy of clusters. The Online DivisiveAgglomerative Clustering (ODAC) system is an incremental implementation of divisive analysis clustering, using the correlation between timeseries as similarity measure. The system tests existing clusters by descending order of diameters, looking for a possible binary split. If no cluster deserves division, then the system searches for possible aggregation of clusters. At each time step, only one splitting or one aggregation might occur. Main features include a splitting criteria supported by the Hoeffding bound, a stopping criteria based on the divisive coefficient and an agglomerative phase which decreases the number of unneeded clusters, also based on the divisive coefficient which measures the amount of divisive structure found. Preliminary results show competitive performance on clustering time-series when compared to a simple batch divisive analysis clustering algorithm.
منابع مشابه
Choosing the Best Hierarchical Clustering Technique Based on Principal Components Analysis for Suspended Sediment Load Estimation
1- INTRODUCTION The assessment of watershed sediment load is necessary for controling soil erosion and reducing the potential of sediment production. Different estimates of sediment amounts along with the lack of long-term measurements limits the accessibility to reliable data series of erosion rate and sediment yield. Therefore, the observed data of suspended sediment load could be used to ...
متن کاملODAC: Hierarchical Clustering of Time Series Data Streams
This paper presents a time series whole clustering system that incrementally constructs a tree-like hierarchy of clusters, using a top-down strategy. The Online Divisive-Agglomerative Clustering (ODAC) system uses a correlation-based dissimilarity measure between time series over a data stream and possesses an agglomerative phase to enhance a dynamic behavior capable of concept drift detection....
متن کاملFuzzy clustering of time series data: A particle swarm optimization approach
With rapid development in information gathering technologies and access to large amounts of data, we always require methods for data analyzing and extracting useful information from large raw dataset and data mining is an important method for solving this problem. Clustering analysis as the most commonly used function of data mining, has attracted many researchers in computer science. Because o...
متن کاملAnytime Concurrent Clustering of Multiple Streams with an Indexing Tree
With the advancement of data generation technologies such as sensor networks, multiple data streams are continuously generated. Clustering multiple data streams is challenging as the requirement of clustering at anytime becomes more critical. We aim to cluster multiple data streams concurrently and in this paper we report our work in progress. ClusTree is an anytime clustering algorithm for a s...
متن کاملA Hybrid Time Series Clustering Method Based on Fuzzy C-Means Algorithm: An Agreement Based Clustering Approach
In recent years, the advancement of information gathering technologies such as GPS and GSM networks have led to huge complex datasets such as time series and trajectories. As a result it is essential to use appropriate methods to analyze the produced large raw datasets. Extracting useful information from large data sets has always been one of the most important challenges in different sciences,...
متن کامل