Academic Paper summary
Online data stream mining contribute to growth on real-world application, such as web click-stream analysis, network traffic monitoring, intrusion detection or credit card fraud detection. The main objective of online data stream mining is to define new mining model and merge into set of built-in & user defined algorithms. It is based on three main technical advantages: new construct and synoptic data structures, fast & light library of mining algorithms which is integrated and stream friendly, and full support for Mining Model Definition Language (MMDL) which allows users to set defined new mining algorithms as tasks and flows. This system goes beyond the previous solutions for more even static mining by providing an extensible workbench for mining, which bring several advantages.
Text available at:
https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=4734034
Comments
Post a Comment