The responsible development of Automated Machine Learning technology in REDD+

Autor: Van Der Zande, David
Přispěvatelé: Dr. Sanneke Kloppenburg
Jazyk: angličtina
Rok vydání: 2021
Předmět:
DOI: 10.5281/zenodo.5902770
Popis: The REDD+ MRV processes depend on technology in several forms, such as the collection of satellite imagery, the development and use of specialised deforestation detection software and the use of mobile devices to collect ground-samples. In this thesis, the use of a new form of data analysis was researched with standardised software development methods for the creation of new innovations in software. A prototype application was developed to showcase the possibilities of such an innovative technology(called AutoML)in REDD+ MRV processes. With the use of value-sensitive design and responsible innovation software development methods, the development and potential use of machine learning for the use of REDD+ MRV processes was analysed with a prototype application, existing scientific literature and expert interviews. The expert interviews were conducted with experts in machine learning and REDD+.
Databáze: OpenAIRE