Advanced Semiconductor Classifiers Using Machine Learning Techniques

Due to the greater dimensional faults that appeared on the wafers earlier in the industry's history, human operators were able to execute inspection activities manually using optical microscopes. The fabrication of semiconductors is a sector that is continually expanding and becoming more significant. The STATISTA website estimates that the worldwide semiconductor sector generated roughly 429 billion USD in revenue in 2019. Testing semiconductors is a crucial step in the production process, especially as the complexity of integrated circuit (IC) designs and the competitive pressure on the market rise. An innovative method to perform advanced semiconductor classification using logistic regression and a random forest classifier is proposed. Semiconductors are found in practically all of the electronics we use on a daily basis. The proposed approach is a unique method in respect of semiconductor testing strategies. Thus, the increased number of test types of devices can significantly increase the cost of manufacturing a single semiconductor chip. This work provides an examination on how to perform automated testing using machine learning techniques.