Measuring and Modeling Food Losses

Abstract Within the context of Sustainable Development Goals, progress towards Target 12.3 can be measured and monitored with Food Loss Index. A major challenge is lack data, which dictated many methodology decisions. Therefore, objective this work to present a possible improvement modeling approach used by Agricultural Organization in estimating annual percentage food losses country commodity. Our proposal combines robust statistical techniques strict adherence rules official statistics. In particular, case study focuses on cereal crops, currently have highest (yet incomplete) data coverage allow for more ambitious choices. Cereal available 66 countries 14 different commodities from 1991 2014. We use loss as response variable, expressed over production, The estimation twofold: it aims at selecting most important factors explaining worldwide, comparing two Bayesian model selection approaches, then predicting Beta regression fully framework.