Mining unusual data over data streams
Wei Xia, Wei Zhang
Abstract
Wei Xia, Wei Zhang
Abstract
It is very important in many fields to mining the unusual data over the data streams. I propose algorithm MUD which are able to quickly and accurately find unusual data from the data streams. Recent works on sample method are not suitable for find the unusual data element from the data streams. In this paper, sample algorithm based on dissimilarity matrix is proposed. The extracted data element from data streams build trend of the unusual data with linear regression model. Experiments show that MUD is effective and the model works perfectly.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
It is very important in many fields to mining the unusual data over the data streams. I propose algorithm MUD which are able to quickly and accurately find unusual data from the data streams. Recent works on sample method are not suitable for find the unusual data element from the data streams. In this paper, sample algorithm based on dissimilarity matrix is proposed. The extracted data element from data streams build trend of the unusual data with linear regression model. Experiments show that MUD is effective and the model works perfectly.
Key concepts: Data stream mining, STREAMS, Data mining, Computer science, Sample (material), Data modeling, Data Matrix, Database