Title :
Discovering frequent serial episodes in symbolic sequences for rainfall dataset
Author :
Ahmed, AlMahdi ; Abu Bakar, Azuraliza ; Hamdan, Abdul Razak ; Abdullah, Sharifah Mastura Syed ; Jaafar, Othman
Author_Institution :
Center for Artificial Intell., Univ. Kebangsaan Malaysia, Bangi, Malaysia
Abstract :
Serial episode is a type of temporal frequent pattern in time series. Many different algorithms have been proposed to discover different types of episodes for different applications. In this paper we propose an algorithm for discovering frequent episodes from processed rain fall data. The algorithm is based on three main steps. (1) The rainfall data is first represented in symbolic representation (2) Then numbers of events are detected by applying sliding window for segmentation and CBR for classification. (3)Finally the processed rain fall data is passed through mining phase. Frequent algorithm is used to discover frequent episodes with fixed width. The experiment shows that many frequent episodes with different structure in different years are extracted.
Keywords :
case-based reasoning; data mining; data structures; geophysics computing; pattern classification; rain; time series; CBR; case base reasoning; data mining; frequent algorithm; frequent serial episode discovery; mining phase; rainfall dataset; sliding window; symbolic representation; symbolic sequences; temporal frequent pattern; time series; Algorithm design and analysis; Classification algorithms; Data mining; Diamond-like carbon; Rain; Time series analysis; frequent episodes; serial episodes; time series;
Conference_Titel :
Data Mining and Optimization (DMO), 2012 4th Conference on
Conference_Location :
Langkawi
Print_ISBN :
978-1-4673-2717-6
DOI :
10.1109/DMO.2012.6329809
