The face recognition is the system in which machine learning is used. The face recognition system is the system which is used for the attendance system. The test and training set is prepared in which segmentation is done, feature extraction can be applied and dataset is designed. The test set is the image which needs to recognize by associating with the preparation set. The SVM classifier is applied which can classify test image into recognized or non-recognized. The linear SVM classifier has least accuracy because it classifies data into two classes. In this research, SVM can be replaced with KNN which increase accuracy of Classification and increase reliability of attendance system.