Authors: Liem Tran*, University of Tennessee at Knoxville
Topics: Spatial Analysis & Modeling, Geographic Information Science and Systems
Keywords: Big data, wavelet analysis
Session Type: Paper
Start / End Time: 5:00 PM / 6:40 PM
Room: Roosevelt 5, Marriott, Exhibition Level
Presentation File: No File Uploaded
Big data and the rise of more complex research opportunities have created a need for more advanced analytical methods. While various advanced theory and methods (e.g., on clustering, associations rules, classification, time series analysis, text analysis, etc.) have been explored extensively, scale issues seem still be pitfalls in big data analytics. The paper explores several scale issues associated with big data analytics via case studies using various environmental and socioeconomic datasets. The authors use multi-level discrete wavelet analysis to decompose a dataset into amplitude/frequency spaces simultaneously to understand how “scale” imposes its “signature” at different levels (i.e., scales) of the phenomenon under study. Those “signatures” can be used to avoid scale-related pitfalls and big data analytics.