This is python and MATLAB codes to calculate the midpoints between k-nearest-neighbor data points of a training dataset (midknn) as a validation dataset in regressionfor. Demonstration of SVR hyperparameter optimization using midknn is also included. For more details on midknn, please go to the URL given as follows: http://datachemeng.com/midknn/ (Please translate the website into English using Chrome, if necessary.)

H. Kaneko, K. Funatsu, Criterion for Evaluating the Predictive Ability of Nonlinear Regression Models without Cross-Validation, Journal of Chemical Information and Modeling, 53(9), 2341-2348, 2013. DOI: 10.1021/ci4003766.