Citas bibligráficas
Rodríguez, A., (2017). Modelos predictivos para determinar la calidad de harina de pescado a partir de características de la materia prima [Universidad Nacional de Trujillo]. https://hdl.handle.net/20.500.14414/21884
Rodríguez, A., Modelos predictivos para determinar la calidad de harina de pescado a partir de características de la materia prima []. PE: Universidad Nacional de Trujillo; 2017. https://hdl.handle.net/20.500.14414/21884
@misc{renati/883703,
title = "Modelos predictivos para determinar la calidad de harina de pescado a partir de características de la materia prima",
author = "Rodríguez León, André",
publisher = "Universidad Nacional de Trujillo",
year = "2017"
}
ABSTRACT The fishmeal processing is based on technology that has been developed, with considerable progress and innovations in recent years. Each production has a number of processes of regulation and control, to ensure its integrity from catching fish to the placing on the market of the final product. This work was based on obtaining mathematical models that predict each of the factors influencing the quality of fishmeal depending on the features that entering the raw material, in order to optimize the time needed for decision making for each of the stages. It defined what are the main characteristics that influence during the process and what methods we will use to determine its value. The factors that determine the quality of fish meal were obtained. With the data we proceeded to use the DataFit 2.0 software for mathematical models that conform as much as possible to reality. The model chosen for the fat content (% Fat) was the linear model with intercept presented a coefficient of determination (R2) equal to 0.9523, amount of residual equal to 0, the value of the Durbin-Watson test is equal to 1.8547 and a significance value 9.708x10-07. The model chosen for the total content of volatile nitrogen (% TBVN) was the linear model with intersection which has a coefficient of determination (R2) equal to 0.9735, residual sum equal to 3.2685x10-13, the value of Durbin Watson Test equal to 1.81 and a significance value 3.729x10-21. By subjecting these models to validation based on the standard deviation, it was concluded that its predictive ability is excellent.
Este ítem está sujeto a una licencia Creative Commons Licencia Creative Commons