International Journal of Academic and Applied Research (IJAAR)   Year: 2019 | Volume: 3 | Issue: 9 | Page No.: 1-8 Smoothing Spline Estimator in Multiresponse Nonparametric Regression for Predicting Blood Pressures and Heart Rate Budi Lestari, Fatmawati, I Nyoman Budiantara Abstract: The basic idea of nonparametric regression is to let the data decide which regression function fits the best without imposing any specific form on it. Consequently, nonparametric methods are in general more flexible. They can uncover structure in the data that might otherwise be missed. In the real cases, we are frequently faced the problem in which two or more response variables are observed at several values of the predictor variables, and there are correlations among responses. For example, blood pressures and pulse are observed at several values of body mass index. Multiresponse nonparametric regression model provides powerful tools for modeling the functions which represent association of these variables. Estimating of regression function is the main problem in this model. Smoothing spline estimator has powerful and flexible properties for estimating the regression function. In this paper we discuss theoretically a method to estimate regression function and optimal smoothing parameter of blood pressures and heart rate models based on smoothing spline estimator in multiresponse nonparametric regression. The estimated regression function can be obtained by taking solution of penalized weighted least square optimization by using reproducing kernel Hilbert space approach. Next, we can get the optimal smoothing parameter by minimizing generalized cross validation function. In this research we obtained plots of predicted values of systolic and diastolic blood pressures and heart rate. The results show that patients who have high BMI, their systolic and diastolic blood pressures, and heart rate lead to hight. It means that patients who have overweight and obese categories lead to rentant suffering hypertension.