Internship (m/f/d) - Geospatial Embeddings

Stelle frei: Ab sofort

Beschreibung

Become part of EIFER and drive climate solutions where science meets industry.

EIFER seeks a student for an internship position (m/f/d) in the field of "Geospatial Embeddings".

Start: 01.07.2025
Duration: max. 6 months (compulsory internship)
Weekly working hours: 39.5
Remuneration: 560€ monthly

In this internship, you will apply machine learning methods in a geospatial context to capture and represent the distinct character of places. You assess its effectiveness in acting as a predictor of missing data, as well as develop and evaluate a more advanced embedding model in the context of an existing open-source library.

Relevant papers (recommended to read in order):
- SRAI: Towards Standardization of Geospatial AI (https://arxiv.org/pdf/2310.13098)
- Hex2vec – Context-Aware Embedding H3 Hexagons with OpenStreetMap Tags (https://arxiv.org/pdf/2111.00970)
- GeoVex: Geospatial Vectors with Hexagonal Convolutional Autoencoders (https://openreview.net/pdf?id=7bvWopYY1H)

Responsibilities
- Applying geospatial machine learning model to prediction tasks
- Calculating urban morphology metrics using open geographic data
- Developing a new geospatial embedding model

Requirements
- Currently enrolled as a University student in a master program (preferred study subjects: Data Science, Computer Science, Geographic Information Systems)
- Good understanding of Python and its data science libraries (especially pandas)
- Familiarity with the Pytorch framework
- Basic understanding of machine learning concepts (embeddings, neural networks)
- Experience with git (push, pull, merge)
- Understanding of geographic information systems (GIS) is a plus
- Confident in working in English
- Strong organizational and communication skills.
- Proficiency in Microsoft Office Suite (Word, Excel, PowerPoint).

Benefits
- Innovative Projects: Engage in impactful research.
- Global Collaboration: Work in an international, diverse team.
- Extensive Networks: Access local and international scientific networks.