Brennan Bean

This paper identifies physical properties of an interesterified palm‐based fat (EIEPO) that predict oil binding capacity (OBC). A 100% EIEPO sample, 50% EIEPO sample diluted with 50% soybean oil (SBO), and a 20% EIEPO sample diluted with 80% SBO were used to test how saturation level impacts OBC. All samples were crystallized using either a fast (6.4°C/min) or slow (0.1°C/min) cooling rate as well as with or without the application of high‐intensity ultrasound (HIU; 20 kHz) to generate a wide range of physical properties. Immediately after crystallization, the sample's physical properties, including crystal microstructure, solid fat content (SFC), viscoelasticity (G′, G″, and δ), melting behavior, hardness, and OBC (centrifuge method) were quantified. The samples were then stored for 48 h at 22 and 5°C and the aforementioned physical properties were measured again, with one additional measurement for the samples stored at 5°C—OBC using a filter paper method (OBC p ). The results indicate that OBC can be optimized in a palm‐based fat by modifying the physical properties which was achieved via the processing conditions. Both measurements of OBC were significantly correlated with SFC, hardness, δ, and enthalpy. A model was developed to predict a sample's OBC c using the following dominant variables—SFC, hardness, peak temperature, enthalpy, and the number of crystals. These results suggest that OBC can be predicted using a sample's SFC, hardness, peak temperature, enthalpy, and number of crystals and that SFC, hardness, and enthalpy are main drivers of OBC.

Accurate yield information empowers farmers to adapt, their governments to adopt timely agricultural and food policy interventions, and the markets they supply to prepare for production shifts. Unfortunately, the most representative yield data in the US, provided by the US Department of Agriculture, National Agricultural Statistics Service (USDA-NASS) Surveys, are spatiotemporally patchy and inconsistent. This paper builds a more complete data product by examining the spatiotemporal efficacy of random forests (RF) in predicting county-level yields of corn—the most widely cultivated crop in the US. To meet our objective, we compare RF cross-validated prediction accuracy using several combinations of explanatory variables. We also utilize variable importance measures and partial dependence plots to compare and contextualize how key variables interact with corn yield. Results suggest that RF predicts US corn yields well using a relatively small subset of climate variables along with year and geographical location (RMSE = 17.1 bushels/acre (1.2 tons/hectare)). Of note is the insensitivity of RF prediction accuracy when removing variables traditionally thought to be predictive of yield or variables flagged as important by RF variable importance measures. Understanding what variables are needed to accurately predict corn yields provides a template for applying machine learning approaches to estimate county-level yields for other US crops.