Explore the unique challenges and techniques of mining spatial and spatio-temporal datasets in this comprehensive lecture presented by Shashi Shekhar at the University of Central Florida. Delve into the fundamentals of spatial data mining, including hotspots, spatial clusters, and outliers. Examine predictive models and learn about the relationships between data in spatial data mining. Gain insights into spatial autocorrelation, statistical foundations, and research challenges in spatial statistics. Discover three general approaches to spatial data mining and their applications in location prediction. Investigate open problems in spatial classification, clustering, and association rules, while understanding the importance of spatial-concept and theory-aware patterns.
Mining Spatial and Spatio-Temporal Datasets: Challenges and Approaches
University of Central Florida via YouTube
Overview
Syllabus
Intro
What is Special about Mining Spatial Data?
Why Data Mining?
Spatial Data Mining (SDM)
Hotspots, Spatial Cluster
Complicated Hotspots
Spatial Outliers
Predictive Models
What's NOT Spatial Data Mining
Relationships on Data in Spatial Data Mining
OGC Simple Features
Research Needs for Data
Statistics in Spatial Data Mining
Overview of Statistical Foundation
Spatial Autocorrelation (SA)
Spatial Autocorrelation: Distance-based measure
Illustration of Cross-Correlation
Spatial Slicing
Edge Effect
Research Challenges of Spatial Statistics
Three General Approaches in SDM
Overview of Data Mining Output
Illustrative Application to Location Prediction
Prediction and Trend
Research Needs for Spatial Classification Open Problems
Clustering
Trends Spatial-Concept & Theory-Aware Patterns
Association Rules - An Analogy
Taught by
UCF CRCV
