JURNAL INFORMATIKA UPGRIS Vol. 8, No. 1 JUNI 2022 P/E-ISSN: 2460-4801/2447-6645 1 Decision Support System For Employee Performance Assessment To Determine The Status Of Reward Level Operator And Foreman Using Adaptive Neuro Fuzzy Inference System (Anfis) Sandy Irawan [1] , Judi Prajetno Sugiono [2] [1][2] Information Technology Master Program Institut Sains Dan Teknologi Terpadu – iSTTS, Surabaya, Indonesia E-mail : [1] shan_classic@yahoo.com E-mail : [2] jpsugiono@stts.edu Abstract— Employees are required to have a good work ethic in order to advance their company. This causes many companies to motivate their employees in various ways. The general goal is for better and more stable employee performance so that it benefits the company. Rewards are given to employees who excel and are able to achieve certain targets, this is more effective in motivating employees than punishment so that it can be a source of motivation for employees to work optimally. In giving rewards, sometimes employees do not match the results of their performance and without applying good calculations. For that we need a recommendation system to support employee performance appraisal to get rewards. One of the methods used is the Adaptive Neuro Fuzzy Inference System (ANFIS) method. This method was chosen because it is able to complete employee performance appraisals based on predetermined criteria and is used as a reference in giving rewards. The amount of data obtained and will be used is a number of 537 employee data which will be divided into two data, namely training data which functions as a model of 524 data and test data which functions to test the system of 13 data. The training set uses a regression algorithm to form an employee performance appraisal model. This model is a representation of knowledge that will be used to predict the reward status of operator and foreman level employees. Keywords—employment; candidates; ANFIS; fuzzy I. INTRODUCTION In the world of work there is intense competition, employees are also required to have a good work ethic. This causes many companies to motivate their employees in various ways. The general goal is to make employee performance better and more stable so that it benefits the company. Reward and punishment are things used by HRD in moving employees to work together in the office. This system has long been known in the world of work. Rewards are given to employees who excel and are able to achieve certain targets, while punishment is given to employees who make mistakes. Rewards or gifts are usually in the form of money, but there are also those who provide rewards in the form of awards, promotions and even holidays. And usually this reward is more effective to motivate employees compared to the threat of punishment or punishment. Many companies offer big rewards for their employees after they achieve certain achievements that even exceed their monthly salary. So it is not wrong for the company or HRD team to use reward and punishment, making it a source of motivation for employees to work optimally. The same applies to PT. AAA, which is the largest integrated flour milling company in one location. In one effort to improve the performance of its employees, PT. AAA gives rewards to employees who have good performance. In giving rewards employees sometimes do not match the results of their performance. Employee rewards are often beaten flat, or sometimes the rewards are only given to them, without applying strong calculations. For this reason, it is necessary to have a recommendation system for supporting employee performance appraisal for reward. One method used is to use the Adaptive Neuro Fuzzy Inference System (ANFIS) method. This method was chosen because it is capable of completing employee performance appraisals based on predetermined criteria, which will later be used as a reference in rewarding itself. The results of the employee performance appraisal decision support system for rewarding also determine what percentage of rewards are given to each employee, so that in this study it will be proven that the ANFIS Method will have higher accuracy compared to other methods that will be applied to the decision support system employee performance appraisal recommendations to determine the status of awarding operator and foreman levels at PT. AAA. Some research related to giving rewards to employees, among others, research conducted by Papageorgiou, et al. (2018) in his research, the fuzzy inference system (FIS) and the adaptive neuro-fuzzy inference system (ANFIS) were developed to classify the total quality of apples based on several fruit qualities, namely fruit mass, meat firmness, dissolved solid content and skin color. The FIS model was evaluated in the same farm for sequential three-year data (2005, 2006 and 2007) and showed 83.54%, 92.73% and