| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 1009.44 KB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Machine-learning methodologies are part of the
artificial intelligence approaches with several applications in different fields of science and dimensions of
human life. These techniques appear in the frameworks
of the digital transition, where smart technologies bring
relevant contributions, such as improving the efficiency
of the economic sectors. This is particularly important for
sectors such as agriculture to deal with the challenges
created in the context of climate changes. On the other
hand, machine-learning approaches are not easy to implement, considering the complexity of the algorithms associated. Taking this into account, the main objective of this
research is to present a model to predict fertiliser costs in
the European Union (EU) farms through artificial neural
network analysis. This assessment may provide relevant
information for farmers and policymakers in the current
scenario where the concerns are to identify strategies to
mitigate the environmental impacts, including those from
the agricultural sector and the respective use of chemical
resources. To achieve these objectives, statistical information for the EU agricultural regions from the Farm Accountancy
Data Network was considered for the period 2018–2020. The
findings obtained show relative errors between 0.040 and
0.074 (showing good accuracy) and the importance of the total
utilised agricultural area and the total output to predict the
fertiliser costs.
Description
Keywords
Farm Accountancy Data Network artificial intelligence European Union agricultural regions
Pedagogical Context
Citation
Publisher
De Gruyter