SubjectsSubjects(version: 964)
Course, academic year 2024/2025
   Login via CAS
Machine Learning in Geosciences - MZ370G24
Title: Machine Learning in Geosciences
Czech title: Strojové učení v geovědách
Guaranteed by: Department of Applied Geoinformatics and Cartography (31-370)
Faculty: Faculty of Science
Actual: from 2024
Semester: summer
E-Credits: 5
Examination process: summer s.:
Hours per week, examination: summer s.:2/2, C+Ex [HT]
Capacity: 15
Min. number of students: unlimited
4EU+: no
Virtual mobility / capacity: no
State of the course: taught
Language: English
Explanation: nahrazuje MZ370P41
Note: enabled for web enrollment
Guarantor: Ing. Lukáš Brodský, Ph.D.
Teacher(s): Ing. Lukáš Brodský, Ph.D.
Incompatibility : MZ370P41
Annotation -
Machine learning has become a significant data science tool to explore and analyze the geography data. The objectives of the course is to review basic principles of machine learning, critically assess the algorithms, practically design processing workflows, apply quality control procedures and interpret the results.
The analysis are applied on spatial and spatio-temporal geography data. Students will develop their own scripts to practically use the gained knowledge of machine learning within the geoscience applications.
There are no formal prerequisites, but a good knowledge of Python language is necessary.
Last update: Brodský Lukáš, Ing., Ph.D. (03.01.2023)
Literature -
  • Bishop C. M. (2006): Pattern Recognition and Machine Learning, Springer.
  • Goodfellow, I., Bengio, Y., Courville, A. (2016): Deep Learning, MIT press.
  • Mehryar, M., Afshin, R., Ameet, T. (2012): Foundations of Machine Learning, MIT press.
  • Lary, D. J., Alavi, A. H., Gandomi, A. H., Walker, A. L. (2016): Machine learning in geosciences and remote sensing. Geoscience Frontiers, Elsevier.

Last update: Čábelka Miroslav, Ing. (15.01.2020)
Requirements to the exam

All assigments from labs (credits).

Written exam, min. 50% correct.

Last update: Brodský Lukáš, Ing., Ph.D. (15.03.2023)
Syllabus -
1. Introduction to Machine Lerning (linear and non-linear problems)
2. Historical context (development of ML algoritms since 1950) 
3. Machine Learning landscape (review of ML types, terminology)
4. Basics of linear algebra & Python data science for Machine Learning (Numpy, Matplotlib, Pandas)
5. Fundamental algorithms (LM, KNN, SVM, DT, ANN, Gradient Descent)
6. Model generalization (training and testing error, components of error, model diagnostics, overfitting, underfitting, model generalization strategies)
7. Model regularization (regularization techniques, regularized linear model, Ridge regression, Lasso, Elastic net, regularizing polynomoial model)
8. Machine Learning project (End-to-end project workflow)
9. Ensemble learning (voting classifier, bagging, stacking, random patches, random subspaces, Random Forest, extra trees, boosting, early stopping)
10. Workign with spatial data in Python (Rasterio, Geopandas, vector-raster combination)
11. Geospatial Artificial Intelligence (spatial dependence, spatial non-stationarity, spatiall cross validation)
12. Introduction to Deep Learning (Pytorch tensors, GPU computing, MLP)
13. Convolutional Neural Networks (spatial data in ANN, CNN idea, CNN elements, CNN loss functions, CNN architectures - FCN, UNet, SegNet)
Last update: Čábelka Miroslav, Ing. (12.06.2024)
 
Charles University | Information system of Charles University | http://www.cuni.cz/UKEN-329.html