Examinando por Materia "Stepwise regression"

Mostrando 1 - 2 de 2

A tool for predicting live weight in Huacaya alpacas from southern Peru
(ResearchersLinks Ltd, England, UK, 2026-07-23) Churata Huacani, Roxana; Canaza Cayo, Ali William; Amarilho Silveira, Fernando; Rodriguez Huanca, Francisco Halley; Marchezan Barchet, Fernanda; Fonseca de Freitas, Rilke Tadeu; Núñez Pérez, Herbert Jesús
The aim of this study was to identify the most suitable model for predicting live weight using both original body measurements (BM) and principal component scores (PC), and to assess the relationship between body measurements (BM) and live weight (LW). LW and BM of Huacaya alpacas (n = 117) were collected from the Quimsachata Center of the National Institute of Agricultural Innovation in Peru. The following BM, including LW, withers height (WH), croup height (CH), thoracic circumference (TC), abdominal circumference (AC), cannon-bone length (CL), neck-base circumference (NBC), cannon circumference (CC), tail insertion length (TIL), rump width (RW), tail insertion circumference (TIC), body width (BDW), rump length (RL), forelimb length (FL), body length (BDL), back length (BKL), were taken. Principal component analysis (PCA) was utilized to extract and clarify the correlation between LW and their BMs. Regression equations relating LW to BM and their PCs were computed. Additionally, the root mean squared error (RMSE), Bayesian information criterion (BIC), Akaike information criterion (AIC), and coefficients of multiple determination (R²) were used to assess the models. Four components were extracted from the PCA of BM and LW, which accounted for 67.3% of the total variance. The prediction model for LW, utilizing two PCs, showed the highest R² as well as the lowest BIC, AIC and RMSE values in contrast to models based on the original measurements. The findings suggest that this method is a viable alternative for predicting live weight and may be useful in breeding programs, as well as in the design of management and selection strategies for Huacaya alpacas from Peru.
Morphometric evaluation of guinea pigs (Cavia porcellus) in Southern Peru
(Learning Gate, 2024-07-19) Quispe Condori, Dennis; Huacani Pacori, Ferdynand Marcos; Mamani Paredes, Javier; Mamani Cato, Ruben Herberht
The aim of this study was the morphometric evaluation of guinea pigs in southern Peru. The study was carried out at the Agrarian Experimental Station Illpa (AESI) of the National Institute of Agrarian Innovation (NIAI) in Puno at 3824 meters above sea level. 120 guinea pigs were used (females n = 60 and males n = 60) with an average age of 22 days. The morphometric characteristics evaluated were: body weight (BW), chest circumference (CC), abdominal perimeter (AP), neck perimeter (NP), head length (HL) and head width (HW). To evaluate the effect of sex on morphometric characteristics, a completely randomized design was used. To determine the equation that best predicts body weight, stepwise regression was used, and correlations between morphometric characteristics were obtained using Pearson's correlation. The results show that the sex factor does not significantly influence BW, CC, AP, NP, HL, and HW (p≥0.05); likewise, the equation that best predicts the body weight of the guinea pigs was: BW = -530.50 + 21.98(CC) + 12.72(AP) + 10.16(NP) + 57.23(HW), with R2 = 84%. Pearson correlations between morphometric characteristics were of high magnitude, positive, and statistically significant (p<0.001). It is concluded that in conditions of the Peruvian highlands, the sex factor does not influence the morphometric characteristics. It is also possible to predict body weight from CC, AP, NP, and HW, and the correlations were high and positive.