Title	Perceptually important point identification for big data analytics : performance analysis and applications
Authors	Hung, Ying Kit
Advisors	Chung, Fu Lai Korris
Degree	Master
Issue Date	2017
Absctract	The concept of the Perceptually Important Point (PIP) identification process is introduced in 2001 in the field of times data mining. This process is originally work for financial time series pattern matching and it is then found suitable for time series dimensionality reduction and representation. The PIP strength is that it is able to preserve the overall shape of the time series by identifying the salient points. With the rise of Big Data, time series data contributes a major proportion, especially on the data which generates by sensors in the Internet of Things (IoT) environment. According to the nature of PIP identification and the successful cases, it is worth to further explore the opportunity to apply PIP in time series “Big Data”. However, the performance of PIP identification is always considered as the limitation when dealing with “Big” time series data. In this dissertation, performance of PIP identification, which always limits the usability of this process, is studied in detail. Improvement algorithms on both algorithmic level and distributed environment are proposed and evaluated. Significant improvement in terms of speed is obtained by these improvement algorithms. In addition, the applications of PIP are reviewed. Then, the usability of PIP for deep learning is evaluated.
Subjects	Data mining Big data Hong Kong Polytechnic University -- Dissertations
Publisher	Hong Kong Polytechnic University
Rights	All rights reserved.