Računalniška delavnica – Strojno učenje na področju opazovanja Zemlje
Opis delavnice: Posvetili se bomo razvrščanju talnih površin z večspektralnimi satelitskimi slikami. Bolj natančno, seznanili se bomo z namensko knjižnico eo-learn in njeno uporabo v procesiranju satelitskih podatkov, pripravili zgledno podatkovno zbirko na manjšem območju in vzpostavili zaporedje procesov, ki je zlahka razširljivo na večje območje, in torej praktično aplikacijo strojnega učenja.
Predznanje: Znanje programiranja v jeziku Python.
Priprava: Napotki za potrebno namestitev so na voljo na https://github.com/JernejPuc/eo-workshop. Ker morebitnih težav ne bomo mogli reševati osebno, je priporočeno, da sledite napotkom in sami predhodno preverite delovanje.
The “Machine learning in the field of Earth observation” workshop will take place on Wednesday, May 6th, 2020, at 15.15 CET through the Zoom video conferencing application (link will be announced later). In case of foreign students, the workshop can be conducted in English.
Description: We will consider the problem of land cover classification with multispectral satellite imagery. Specifically, we will get familiar with the eo-learn framework and how it can be used to process satellite image data, then use it to prepare a small data set and to build a pipeline that can be easily scaled, leading to a practical machine learning application.
Prerequisites: Knowing how to program in Python.
Preparation: The set-up guide is available on https://github.com/JernejPuc/eo-workshop. As we won't be able to resolve issues in person, it is recommended that you follow the instructions and make sure that everything works in advance.
Prijave / Sign-up: https://forms.gle/kYwLJCNVR4VUjrD87
Se vidimo!
