Open AccessPre-processed Hierarchical Clustering for Time Series Data Streams
Author(s) -
Dr.V. Kavitha*,
Dr.A.V.Senthil Kumar,
Dr.N. Revathy,
Mr.C.Daniel Nesa Kumar,
Mrs.P. Hemashree
Publication year - 2019
Publication title -
international journal of recent technology and engineering
Language(s) - English
Resource type - Journals
ISSN - 2277-3878
DOI - 10.35940/ijrte.c3961.098319
Subject(s) - computer science , data stream mining , process (computing) , cluster analysis , task (project management) , signal (programming language) , data stream , artificial intelligence , data mining , pattern recognition (psychology) , feature (linguistics) , transformation (genetics) , feature selection , signal processing , time series , series (stratigraphy) , key (lock) , selection (genetic algorithm) , machine learning , engineering , digital signal processing , philosophy , computer security , systems engineering , linguistics , chemistry , biology , operating system , telecommunications , paleontology , biochemistry , programming language , computer hardware , gene
The behaviour of the human body is based on the signals of chemical, electrical origin. These signals afford information that may not be directly perceptible but some information is hidden in the structure of the signal. These hidden signal information has to be translated in some way before the signals can be given useful analysis. The transformation of human body signals has been discovered useful in explaining and identifying various pathological conditions. The process of transformation is comfortable to perform since involves a limited manual effort like visual investigation of the signal generated as a result. In spite of these signals with their complexity is often considered and consequently biomedical signal processing has become an essential task for extracting significant clinical information hidden from the original signal[1]. Time series data streams constitute numerous dimensions and noisy features. Therefore, detecting the original clusters in high dimensional noisy features time series data stream is a dispute task. The challenging task involved in time series data stream are noisy and high dimensional. The existing technique is incapable of handling noisy high dimensional data stream. The most important key objective of this research part is to develop a novel pre-processing feature selection technique for discarding the noisy data is a vital successful process. Therefore this technique achieves minimum time complexity. Pre-processing feature selection is an established technique to deal with the time series data stream with noisy and high dimensional [3]. Furthermore, this innovative feature selection approach is boost up the cluster process without noisy and also it accomplishes the quality clusters with minimal time interval.