User satisfaction with machine learning as a data analysis method in agricultural research

Abstract

Abstract Machine learning has potential use in the understanding of information hidden in large datasets, but little is known about user perceptions of the use of the technology. In this study, datasets were solicited from agricultural researchers and processed using a machine learning workbench. The results were reported to the researchers, and then interviews were conducted with some of them to determine their perceptions about the use of machine learning as an additional analysis technique to traditional statistical analysis. A number of themes about their satisfaction with this technique were constructed from the interview transcripts, which generally indicate that machine learning may be able to contribute to analysis and understanding of these kinds of datasets.