Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Mimos BerhadfiledCriticalMimos Berhad
Priority to MYPI2015001665ApriorityCriticalpatent/MY178650A/en
Publication of MY178650ApublicationCriticalpatent/MY178650A/en
Information Retrieval, Db Structures And Fs Structures Therefor
(AREA)
Abstract
The present invention relates to systems and methods for detecting abnormal instances in a dataset for a given time granularity (e.g., daily, weekly, monthly). The system comprises a pre-processing module (202) that receives a plurality of raw transactions and.generates one or more time instance, Tl datasets based on a defined time granularity and a data restructuring and re-clustering module (212) that.selects at least one optimal cluster model by iteratively clustering the Tl datasets and re-structuring normal transactions of the clustered Tl dataset with a subsequent Tl dataset of one or more Tl datasets using multiple cluster models. The optimal cluster model is then used for processing new transactions for detecting anomalies by running a new Tl dataset generated for the new transactions on the optimal cluster model for clustering them into an abnormal cluster having highest number of abnormal transactions and at least one normal cluster having normal transactions.
MYPI2015001665A2015-06-292015-06-29System and method for data anomaly detection
MY178650A
(en)
Computer-implemented tools and methods for extracting information about the structure of a large computer software system, exploring its structure, discovering problems in its design, and enabling refactoring