Nonparametric Regression Model Based on Estimator of Kernel Local Polynomial on the Jackfruit Seed Growth

Regression analysis is one statistical data analysis techniques in the most widely used to determine the relationship between the predictor variable (X) with the response variable (Y). In regression analysis, the regression curve can be approximated by nonparametric regression model. Nonparametric regression model that the observations were made of n subjects are mutually independent at a certain point so that it follows the structure of cross-section data. In this research will be assessed on nonparametric regression model based on local polynomial estimator kernel on the growth of seedling of jackfruit. Data sources and research variable used in this research is primary data from Agriculture faculty of KH. A. Wahab Hasbullah University at Jombang. The data on the growth of plant height (X) is effected plant age (Y). Analysis method are estimates a nonparametric regression model based on the local polynomial kernel estimator and the result of estimation obtained is applied to analyze the influence of plant height variable of jackfruit seedlings. Local kernel polynomial estimators can be obtained by WLS (Weighted Least Square) optimization. As for estimating the smoothing parameter (bandwidth) using GCV (Generalized Cross Validation).Modeling done on granting Urea and plant growth regulator Auxin, the provision of Urea has a value of MSE = 1.58133 and the coefficient of determination (R2) = 80.16%. The PGR auxin has a value of MSE = 1.267782 and the coefficient of determination (R2) = 78.59%. Based on the result of the research, it is estimated that the growth of jackfruit seedlings using the Local Polynomial Kernel method of estimation error rate with the treatment of Urea fertilizer is higher by 19.85% compared to the treatment of PGR Auxin. This means that the estimated growth of seedlings of jackfruit on providing the best estimate of PGR Auxin results