01522nam a2200229 a 450000100080000000500110000800800410001910000170006024501140007726000490019152008680024065000200110865000190112865000090114765000230115665000140117965000270119365000100122065300200123070000200125070000220127020717502017-06-29       2017    bl uuuu     u00u1 u    #d1 aOLIVEIRA, G.  aComparing priors in bayesian logistic regression for sensorial classification of rice.h[electronic resource]  aIn: SAS GLOBAL FORUM 2017, Orlando, FLc2017  aThe present study use Proc MCMC to estimate rice stickiness (sticky or loose) in binomial logistic regression. Rice quality can be primarily assessed by evaluating its texture after cooking. The classical sensory evaluation is expensive and a time-consuming method since it requires training, capability and availability of people. Therefore, the present study investigated Bayesian binomial logistic models to replace sensory evaluation of stickiness by analyzing the relationship between sensory and principal components of viscosity measurements of rice. Proc MCMC was used to produce models based on different priors, as (1) noninformative prior; (2) default prior (Gelman et al., 2008); (3) prior based on odds ratio (Sullivan and Greenland, 2012) and (4) power priors (Ibrahim and Chen, 2000). SAS MCMC showed to be easy to implement and to compare results.  aBayesian theory  aGlutinous rice  aRice  aSensory evaluation  aViscosity  aAnálise organoleptica  aArroz  aClassificação1 aVON BORRIES, G.1 aBASSINELLO, P. Z.