EvolveCluster: an evolutionary clustering algorithm for streaming data

Data has become an integral part of our society in the past years, arriving faster and in larger quantities than before. Traditional clustering algorithms rely on the availability of entire datasets to model them correctly and efficiently. Such requirements are not possible in the data stream clustering scenario, where data arrives and needs to be analyzed continuously. This paper proposes a novel evolutionary clustering algorithm, entitled EvolveCluster, capable of modeling evolving data streams. We compare EvolveCluster against two other evolutionary clustering algorithms, PivotBiCluster and Split-Merge Evolutionary Clustering, by conducting experiments on three different datasets. Furthermore, we perform additional experiments on EvolveCluster to further evaluate its capabilities on clustering evolving data streams. Our results show that EvolveCluster manages to capture evolving data stream behaviors and adapts accordingly.

EMM-CLODS: An Effective Microcluster and Minimal Pruning CLustering-Based Technique for Detecting Outliers in Data Streams

Detecting outliers in data streams is a challenging problem since, in a data stream scenario, scanning the data multiple times is unfeasible, and the incoming streaming data keep evolving. Over the years, a common approach to outlier detection is using clustering-based methods, but these methods have inherent challenges and drawbacks. These include to effectively cluster sparse data points which has to do with the quality of clustering methods, dealing with continuous fast-incoming data streams, high memory and time consumption, and lack of high outlier detection accuracy. This paper aims at proposing an effective clustering-based approach to detect outliers in evolving data streams. We propose a new method called Effective Microcluster and Minimal pruning CLustering-based method for Outlier detection in Data Streams (EMM-CLODS). It is a clustering-based outlier detection approach that detects outliers in evolving data streams by first applying microclustering technique to cluster dense data points and effectively handle objects within a sliding window according to the relevance of their status to their respective neighbors or position. The analysis from our experimental studies on both synthetic and real-world datasets shows that the technique performs well with minimal memory and time consumption when compared to the other baseline algorithms, making it a very promising technique in dealing with outlier detection problems in data streams.