At this time many researchers use datasets for research, in the dataset there is a lot of important information provided to make it easier for researchers to process the data, but there are obstacles when trying to process the data, namely some data values are lost, or even damaged (incompatible) with other values. The number of attributes and data samples in the dataset is unlimited which makes it difficult for researchers to find important information based on the goals of each researcher. The missing or damaged values are known as missing values and this often happens when researchers process data taken from datasets. When researchers classify a dataset, of course it is difficult to classify because if there is missing data, the classification process cannot be carried out completely. Of course this will greatly affect the results obtained and also the results of the accuracy of the process. The problem of missing data (missing value) must be resolved, one way to solve this problem is by using the K-Nearest Neighbor Imputation (K-NNI) algorithm. The KNNI algorithm is one of the data mining algorithms that can restore the missing value problem by performing pattern recognition of the nearest data from the missing data. The final result after applying KNNI in fixing the missing value is to obtain a weight value for each K closest observation of 51.6161 on K5. Datasets that experience missing values have been adjusted based on the weight values and the implementation of the KNNI algorithm with the rapidminer application. By returning data using the KNNI algorithm, it is hoped that it can help solve these problems so that the results obtained from the classification process are more accurate and the results of greater accuracy