[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-118869-en":3,"doc-seo-118869-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},118869,8796095360427,"Lucas Martin","https://ap-avatar.wpscdn.com/davatar_994ba38a5ba835b3df7d355c54d3ed8d",8,"Research & Report","Comparison of Classification Machine Learning Models for Production Flow Analysis in a Semiconductor Fab - Research Report","A semiconductor fab involves complex wafer-lot movements across machines and workstations, making throughput performance dependent on batching, allocation, scheduling, processing times, and queue status. This study applies multiple machine learning classification models to predict weekly throughput levels using real-time process and utilization indicators collected via a digital-twin concept. Model accuracies are compared to identify the best-performing approach, with random forest achieving the highest accuracy above 97%. The work extends earlier research by evaluating additional model options and proposing a more comprehensive feature-selection and model-comparison framework.","Comparison of Classification Machine Learning Models for Production Flow Analysis in a Semiconductor Fab  \nIvan Kristianto Singgih 1,2,3,*, Stefanus Soegiharto 1 , Arida Ferti Syafiandini4,5  \n1 Department of Industrial Engineering, University of Surabaya, Surabaya, Indonesia  \n2 The Indonesian Researcher Association in South Korea (APIK), Seoul, 07342, South Korea  \n3 Kolaborasi Riset dan Inovasi Industri Kecerdasan Artifisial (KORIKA), Jakarta, Indonesia  \n4 Department of Library and Information Science, Yonsei University, Seoul, South Korea  \n5 Research Center for Computing, National Research and Innovation Agency, Indonesia (BRIN), Cibinong, Indonesia  \n*Corresponding author  \nAbstract. A semiconductor fab has complex wafer lot movements between machines and workstations. To ensure a smooth flow of the wafer lots, the system must be observed appropriately. Observation of such a complicated system is possible using machine learning. In this study, various machine learning techniques are applied to predict the semiconductor fab’s throughput when considering wafer lot processing and queuing status at the machines and the machine utilization. The accuracies of the models are compared. It is shown that the random forest model obtained the best accuracy of more than 97% . Compared with the previous study, this study considers more models to allow a more comprehensive evaluation. The findings are important for providing suggestions on machine learning model selection for predicting the output of a semiconductor fab.  \nKeywords: Semiconductor Fab, Classification, Prediction, Machine Learning,  \nModel Evaluation.  \n1 Introduction  \nSemiconductor fab has a complex environment due to the re-entrants of wafers to workstations and the usage of parallel machines [1] . The movement of wafers in a semiconductor fab (Intel minifab) is illustrated in Figure 1. Before being processed at workstation 1, the wafer lots must be grouped into batches first. Please refer to Singgih [1] to obtain more details on the simulation used for the data collection, the required processing times on the machines, and the arrival schedule of the wafer lots.  \n© The Author(s) 2023  \nM. Hartono et al. (eds.), Proceedings of the 4th International Conference on Informatics, Technology and Engineering 2023 (InCITE 2023), Atlantis Highlights in Engineering 21, [https://doi.org/10.2991/978-94-6463-288-0_24](https://doi.org/10.2991/978-94-6463-288-0_24)  \nComparison of Classification Machine 269  \nSinggih [1] has shown that some classification machine learning models could be used to predict the system’s weekly throughput when considering various product and machine-related information, e.g., the number of processed wafer lots on each machine, number of wafer lots on the machine queues, and machine utilization.  \nFig. 1. Wafer flow in Intel semiconductor minifab.  \nFig. 2. The digital twin framework for the semiconductor process analysis.  \nThe system proposed by Singgih [1] was a digital twin (Figure 2) . In this digital twin system, real-time data related to the wafer lot processing and machine status can be collected using IoT sensors placed on the machines and their queues. This wafer lot processing and machine status show how well the operation was optimized, e.g., using batching, allocation, and scheduling decisions. In other words, when good optimization decisions are made, the wafer lot flow would be smooth, e.g., all machines are highly utilized, less queuing times, etc., and the weekly target throughput would be satisfied. By observing the production status information as input data, and the throughput satisfaction level (e.g., low and high) as the target (output) each week, Singgih [1] stated that the relationships between those input and output data could be identified. Using the same framework for finding the relationships between the input  \n270 I. K. Singgih et al.  \nand output data, the importance of each input data could also be observed, e.g., by iter","cbCaitMMSC0aVPHl","https://ap.wps.com/l/cbCaitMMSC0aVPHl","pdf",1549300,1,9,"English","en",105,"# Introduction\n## Digital twin framework and data inputs\n## Study contributions\n# Solution Methodology\n## Considered model families\n# Numerical experiments and analysis\n# Conclusion","[{\"question\":\"What is the main goal of this study in semiconductor production analysis?\",\"answer\":\"To predict the semiconductor fab’s weekly throughput satisfaction level by applying and comparing multiple classification machine learning models using machine processing and queue-status information.\"},{\"question\":\"How does the study obtain input data for machine learning?\",\"answer\":\"It relies on a digital twin approach where real-time wafer-lot processing and machine status are collected using IoT sensors for machine utilization and queue conditions.\"},{\"question\":\"Which model achieved the best prediction accuracy?\",\"answer\":\"The random forest model achieved the best accuracy, exceeding 97%.\"}]","Comparison of Classification Machine Learning Models for Production Flow Analysis in a Semiconductor Fab - Research Report | PDF",1785720704,23,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"comparison-of-classification-machine-learning-models-for-production-flow-analysis-in-a-semiconductor-fab-research-report","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/comparison-of-classification-machine-learning-models-for-production-flow-analysis-in-a-semiconductor-fab-research-report/118869/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-03",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What is the main goal of this study in semiconductor production analysis?","Question",{"text":75,"@type":76},"To predict the semiconductor fab’s weekly throughput satisfaction level by applying and comparing multiple classification machine learning models using machine processing and queue-status information.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the study obtain input data for machine learning?",{"text":80,"@type":76},"It relies on a digital twin approach where real-time wafer-lot processing and machine status are collected using IoT sensors for machine utilization and queue conditions.",{"name":82,"@type":73,"acceptedAnswer":83},"Which model achieved the best prediction accuracy?",{"text":84,"@type":76},"The random forest model achieved the best accuracy, exceeding 97%.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,127,130,134],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":21,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]