Practical machine learning for spatial data - Practical machine learning for spatial data - Training
|Date:||30.03.2020 9:00 - 01.04.2020 16:15|
|Location details:||The event is organised at the CSC Training Facilities located in the premises of CSC at Keilaranta 14, Espoo, Finland. The best way to reach us is by public transportation; more detailed travel tips are available.|
|lecturers:|| Mats Sjöberg |
|The fee covers all materials, lunches as well as morning and afternoon coffees.|
Payment can be made with electronic invoicing, credit card, or direct bank transfer. Note that for electronic invoicing you need the operator and e-invoicing address (OVT code) of your organization. Please also note that invoice reference is needed for electronic invoicing in your organization, so please have this available when registering.
This course gives a practical introduction to machine learning for spatial data, both to shallow learning and deep learning models, especially convolutional neural networks (CNN).
The course consists of lectures and hands-on exercises. OrfeoToolBox and scikit-learn will be used for the shallow learning exercises on local PCs. keras, (pytorch) and solaris will be used for deep learning exercises on Puhti-AI.
If you have not worked with Puhti or Taito before, please attend the optional Puhti intro at the end of second day.
After the course the participants should have the skills and knowledge needed to begin applying machine learning and deep learning for different tasks and utilizing the GPU resources available at CSC for training and deploying their own neural networks.
- Basics of geoinformatics
- Basics of Python. The course will include a fair amount of reading and writing Python code, so you should be able to follow Python syntax. If you need to refresh your Python skills you can go through the materials of Helsinki University GeoPython course.
- very basic Linux commands: cd, ls, mv, cp, rm, chmod, less, tail, echo, mkdir, pwd. If unfamiliar take a look for example at LinuxSurvival first two modules.
The course is similart the Practical machine learning for spatial data course kept in autumn 2019.
|9:00-10:15||Lecture: Introduction to machine learning|
|10:30-12:00||Lecture: Introduction to exercises, preparing spatial data for machine learning |
Exercise 1: Preparing vector data for regression
Exercise 2: Preparing raster data and labels for clustering and classification
|13:00-14:30||Lecture: Shallow machine learning models|
|14:45-16:15||Lecture: Software supporting machine learning of spatial data |
Exercise 3: Using OrfeoToolBox for machine learning
|9:00-10:15||scikit-learn exercises: |
Exercise 4: Shallow regression
Exercise 5: Image segmentation using k-means
Exercise 6: Image classification using shallow classifiers, grid search
|10:30-12:00||Exercise 8: Shallow learning exercise in pairs|
|13:00-14:30||Lecture: Introduction to deep learning models |
Lecture: Fully connected neural networks
Lecture: GPUs and batch jobs
|14:45-16:15||Optional: Puhti intro and hands-on|
|9:00-10:30||Exercise 9: Fully connected regressor |
Exercise 10: Fully connected classifier
Exercise 11: Fully connected classifier in pairs
|10:45-12:00||Lecture: Convolutional neural networks (CNN)|
|13:00-14:30||Exercise 12: CNN based image segmentation|
|14:45-16:15||Exercise 13: CNN based image segmentation in pairs|