Mississippi State University Harnessing Machine Learning to Estimate Aquaculture Production and Value Chain Performance in Bangladesh Annual Report April to September 2020


Views
0% 0
Downloads
0 0%

CC-BY-4.0

The Harnessing Machine Learning for Better Aquaculture project began operations on April 1, 2020. The project is implemented by Michigan State University (MSU) in partnership with Bangladesh Agricultural University (BAU), WorldFish (Bangladesh office) and the International Center for Tropical Agriculture (CIAT). The project has three goals First, identify emerging technologies and innovative practices in aquaculture value chains and pilot digital extension approaches that accelerate their adoption. Second, use machine learning to automate extraction of data on ponds from satellite images and integrate with georeferenced survey data to accurately estimate fish production, economic value, and employment. Third, build organizational and individual capacity in Bangladesh for conducting rigorous research on socioeconomic and spatial dimensions of aquaculture. The project is comprised of three components that feed into these two goals: (1) Surveys; (2) Remote sensing; (3) Capacity building. Component 1 will survey a sample of 1100 hatcheries, feed suppliers, farmers, and fish traders. Component 2 will utilize machine learning to extract and analyze data on fishponds from satellite images to facilitate development of an interactive online data visualization tool utilizing data from component 1. Component 3 is dedicated to formal training and outreach that builds individual, organizational and societal capacity. Good progress was made during the first two quarters of the project (1 April 2020 - 30 September 2020) towards activities under components 1 and 2. Activities oriented to component 3 will be initiated later on in the project cycle.

Citation

Ben Belton. (30/9/2020). Mississippi State University Harnessing Machine Learning to Estimate Aquaculture Production and Value Chain Performance in Bangladesh Annual Report April to September 2020.

DOI

Other URI

Date available

Publisher

Country(ies)

Language(s)

Keyword(s)