<![CDATA[Productivity is expected to support the company's goal of achieving high profits. Technician performance assessment still relies on subjective assessment from the team leader, causing difficulties in objective evaluation. This research aims to improve the appraisal of technician performance at PT Telkom by proposing using the k-nearest Neighbor (k-NN) method as a classification method. The k-NN method was chosen because of its robustness to training data and effectiveness on large training data. The classification process includes finding the distance between two points using the Euclidean equation, sorting the lowest value as the nearest neighbor, calculating the number of neighbors based on the 5 (five) nearest neighbors for the classification of "Dissatisfied", "Quite Satisfied,", "Satisfied", and "Very Satisfied". Based on calculations carried out using a confusion matrix, this algorithm has an accuracy rate of 74%, precision of 80%, and recall of 80%, so it can be recommended to PT Telkom to know the actions to be taken, whether a technician can be retained or not in the company]]>
Copyrights © 2023
