Paper Title
An inteligent System for Detecting Defects in Products

Abstract
In this article, we conducted an investigation into the intelligent identification of defects occurring in the injection molding process. The implementation of an intelligent system for defect detection in products brings significant benefits and advancements to quality control and manufacturing procedures. We outlined the various types of defects targeted for detection and the input variables employed in the intelligent algorithms. Subsequently, we presented the construction of our intelligent system. Additionally, we performed a comparison among multiple intelligent algorithms to determine the most accurate classifier. "K-Nearest Neighbors" emerged as the top performer, achieving an accuracy of over 96% for all defect types, closely followed by "Decision Tree" with an accuracy exceeding 95%. Keywords - Burr; Not Complete; Dark Spot; Defect Detection; Defect Type; AI; IoT