Ghaffarian, Saman; de Mey, Yann; Valente, João; van der Voort, Mariska; Tekinerdogan, Bedir; (2023) Deep learning for agricultural risk management: Achievements and challenges. In: Tekinerdogan, B and Catal, C and Alskaif, T and Akşit, M and Hurst, W, (eds.) Management and Engineering of Critical Infrastructures. (pp. 307-333). Elsevier: London, UK.

Text (Chapter 14) Deep Learning for Agricultural Risk Management.pdf - Accepted Version Access restricted to UCL open access staff Download (566kB)

Abstract

Farms face various risks such as uncertainties in the natural growth process, stiffened financing sources, volatile prices, lack of market access, unpredictable changes in farm-related policies and regulations, and farmers’ health problems. Accordingly, farmers have to make decisions to be prepared for such situations or mitigate their impacts and maintain essential functions. Deep learning (DL) has become a state-of-the-art machine learning method for automatic processing of different data types in various science fields, including the agricultural risk management (ARM) domain. DL methods have been employed to support decision making in agriculture through the extraction of effective information from data. The potential of DL in ARM has recently increased with advances in technology and digitalization. This review aims to complement prior reviews by identifying and analyzing state-of-the-art advances in DL-based ARM (DL-ARM), and thus, demonstrate the current achievements and challenges. In particular, a systematic literature review approach is adopted to extract current trends in publications, publishers, used DL methods, applications, lessons learned, weaknesses, and open problems, to formulate recommendations for researchers and field practitioners. The papers were categorized based on the addressed risk types, specific objectives, deep learning algorithms, and data types. In total, 203 papers were retrieved, of which 100 papers were selected based on the defined exclusion criteria for a detailed review. Guidelines are discussed on how DL can further contribute to ARM by shifting focus away from assessing only production risk and leveraging DL to make decisions rather than simply support making them.

Download activity - last month

Export as JSONCSVXML

Download activity - last 12 months

Export as XMLJSONCSV

Downloads by country - last 12 months

Export as XMLJSONCSV

Archive Staff Only

View Item