VGA (Video Graphics Array) is a Video adapter which is very useful for improving the performance and quality of the visual process on a computer, but sometimes there is often a malfunction that cannot be identified the type of damage. The problem is the lack of media to identify the damage that occurs during visual processing. Therefore, the authors created an expert system that can diagnose the type of damage to VGA using the Certainty Factor method as a calculation, using UML modeling as the work process flow of the system on the website, and also equipped with the KNN (K-Nearest Neighbor) algorithm as machine learning. so that it can build an expert system with the PHP programming language MySQL database. The method used in testing is the black box method in testing the system used. The results that can be concluded from this study are; 1) The diagnostic system for detecting damage to the VGA uses the K-Nearest Neighbor Algorithm as machine learning and the Certainty Factor Method as a calculation medium in determining the distance from the type of damage and has suggestions for further actions to deal with and prevent the damage from occurring and also has other possible damage things that are similar to the damage suffered can be accessed quickly and easily to understand, in making scientific research carried out sequentially to facilitate the process, and 2) In addition to diagnosing, there are several additional menus that can be accessed such as the Prediction menu which functions to displays the max and min limits of the temperature of a product, Product Info which functions as a quality product recommendation, and a description that contains a post of details of the damage that can be studied and is expected to help users find solutions to their problems.Keywords:Expert System, PHP, Certainty Factors, Machine learning, K-NN.