2025-03-272025-03-272020OLIVEIRA, Sara Caroline Pacheco de; FERREIRA, Thalia de Oliveira. Predição de parâmetros físico-químicos de tucupi amarelo utilizando espectroscopia no infravermelho próximo (NIRS). 58 f. Trabalho de Conclusão de Curso (Tecnologia de Alimentos) – Universidade do Estado do Pará, Belém, 2020.https://bibliotecadigitaldetcc.uepa.br/handle/riuepa/566Cassava (Manihot esculenta) is a tuberous root of the euphorbia family, with originality from South America and easy to adapt mainly in tropical regions. From it the most diverse products are obtained, among them: the tucupi. Tucupi is used as a main ingredient in the preparations of paraense cuisine. As there is no concern about the provenance, in this work an analysis of the yellow tucupi was carried out to assess its quality and the possibility of predicting its physical-chemical parameters by means of near infrared spectroscopy (NIRS). For this purpose, 65 were analyzed in 10 different neighborhoods in the metropolitan region of Belém – PA, obtaining average values of titratable acidity expressed in lactic acid (0.64±0.23 g / 100 mL), total sugars (0.44±0.46 g / 100 g), total starch (1.41±0.81 g / 100 g), specific weight (0.92±0.01 g / 100 g), pH (3.80±0.30), SST (3.82±0.77 ° Brix) and chromatic parameters of L* (49.06±7.33), a* (-5.24±2.77), b* (33,69±14.89), hº (99.43±4.38) and C* (34.39±14.99). The results obtained were submitted to variation analysis (ANOVA) followed by Tukey's comparative test. It was also possible to make correlations between the investigated parameters by determining the Pearson coefficient (r), which resulted in the comparison of results between Fairs × Supermarkets (F × S), Fairs × Fairs (F × F) and Supermarkets × Supermarkets (S × S). Therefore, in general the samples differed little statistically (P <0.05) mainly for the parameters of acidity, pH, SST and specific weight, as well as for the colorimetric parameters a *, b * and Hº. However, despite both establishments presenting irregularities, the samples obtained at fairs stood out in the parameters of total sugars and total starch and in the chromatic parameters L * and C *. The NIR spectrum was used for predictions of physical-chemical parameters, using the technique of partial least squares (PLS), where the results obtained for predictive calibration and cross-validation were considered unsatisfactory for obtaining the prediction values in tucupi samples (R2> 0.800) and (RPD> 2.0).Língua PortuguesaAbertoTucupiNear infrared spectroscopyAnalysis of varianceCorrelationPredictionTrabalho de Conclusão de Curso (TCC)