Lecture Principles of Spatial Data Mining and Machine Learning

Aim of this Lecture

In this lecture, the students learn how the field of data mining has originated from predictive modeling, the core techniques of unsupervised (clustering) and supervised data mining are introduced (linear/logistic regression, SVM, RF, CNN, etc.) and applied in both a classification and a regression task. Special attention is given to spatial data including relevant algorithms, treatment of ML fitting problems, treatment of feature selection, model construction, model fusion, and data cleaning.

By completing this module, students will be enabled to extract knowledge from spatial and spatio-temporal datasets following techniques from data mining and machine learning including linear models, kNN models, regression models, classification models, decision trees, Support Vector Machines and more. These methods are applied to spatial datasets including satellite images, spatial networks, and geo-social media data. Students get an overview of methods and techniques to explore big geospatial datasets using data mining techniques.

Zoom Information

Note that below information is for convenience of not fully registered students and might be changed or disappear without public notice. Please register (TUMonline / Moodle).

https://tum-conf.zoom.us/j/68519710317

Meeting ID: 685 1971 0317 Passcode: 447436

The timeslot is currently

Friday, 13:15 - 15:30 (CET)


© 2020 M. Werner