Cross-validation based kNN imputation, designed and developed by J. Huang

Title: Cross-validation based K nearest neighbor imputation for software quality datasets: An empirical study Published on The Journal of Systems and Software Authors: J. Huang, etc.

https://www.sciencedirect.com/science/article/abs/pii/S0164121217301516

Please find the paper: 1-s2.0-S0164121217301516-main.pdf

The cvbknni was based on the knn kernel, with cross-validation to optimize the various knn parameters for various missingness in the data. The original code was written in Matlab R2017. (Matlab R2016a - R2018b should be fine)

cvbknni was written as IDknnimpute.m and IDknnimpute3.m, the difference was imputation ordering methods.

if you find anything not right, or missing component, please contact me via hellwalkerdh[at]gmail[dot]com