Introduction to Machine Learning

This course was developed as a MEST Micro-Certificate. MEST Micro-Certificates offer short, self-paced learning opportunities through focused training modules and webinars on specialized technical topics. Participants who successfully complete the program requirements earn a certificate of completion.

These programs are designed for U.S. citizens, including Department of War and other government personnel, government-affiliated employees, industry professionals, college students, and faculty.


Machine learning is a powerful tool for analyzing and making predictions from data. This course introduces the fundamental concepts of machine learning and provides hands-on experience with training and evaluating machine learning models in the context of microelectronic manufacturing problems. The course covers traditional machine learning, such as Nearest Neighbors, LDA, QDA, Naive Bayes, Decision Trees, and Support Vector Machines (SVMs), as well as modern deep learning model fundamentals as far as Convolutional Neural Networks (CNNs). The course also includes multiple case studies of selected applications of machine learning to the microelectronics domain.

Machine learning represents a class of algorithms that assimilate large quantities of data to tune an internal model, then can generalize to data not seen in their given dataset (learning.) In the context of microelectronics, machine learning algorithms can assist with the analysis of large datasets, such as wafer maps, to identify patterns and trends that are difficult to discern with traditional statistical methods. The machine learning methods presented in this course are divided into three modules:

  • Unit 1: Overview of Machine Learning: Establishes the foundational perspective and views for understanding application of machine learning concepts to the domain-specific problems in the microelectronics domain. Introduces fundamental traditional machine learning algorithms, such as Nearest Neighbors, LDA, QDA, Naive Bayes, and Decision Trees.

  • Unit 2: Domain Specific Machine Learning in Microelectronic Field: Focuses on the fitting error vs model complexity tradeoff in the generalization of traditional machine learning approaches, showing how they bridge to modern Deep Learning techniques. Then evaluates case-studies of these ML approaches in the microelectronics domain, both in lecture and homework.

  • Unit 3: Unsupervised Learning in Microelectronic Field: Introduces Unsupervised Learning’s fundamental differences from the supervised techniques covered up to this point, framing the challenges of previous techniques in the pipeline of microelectronic device manufacturing. Then deeply explores representative techniques in unsupervised learning and the risks and benefits of application in microelectronics’ domain-specific context.

  • Unit 4: (Supplemental) Machine Learning in Silicon Data Analytics: Reviews the benefits and challenges of machine learning overall in application within the microelectronic domain with specific research case studies and techniques developed by the UCSB research team supplying the course content.

Format: Online, asynchronous
Platform: nanoHUB Learning Management System
Estimated Time: 15-18 hours
Level: Advanced Undergraduate/Graduate Student
Prerequisites:

  • An understanding of high-level programming (loops, objects, functions, libraries) and basic college-level mathematics (algebra, linear algebra, trigonometry, derivatives) is assumed.

  • Knowledge of the Python programming language and the fundamental units of microelectronics fabrication (wafers, die, lithography) is helpful but not required.

Instructor: Professor Li-C Wang, University of California, Santa Barbara


Course Objectives:

  1. Apply fundamental algorithms such as Nearest Neighbors, LDA, QDA, Naive Bayes, and Decision Trees to microelectronics problems

  2. Analyze the trade-off between fitting error and model complexity in traditional and deep-learning models

  3. Explain the principles and risks of unsupervised learning in the microelectronics manufacturing context

  4. Evaluate case studies of machine-learning applications in silicon data analytics

 

Request Course Enrollment

To enroll in this course, you must have a nanoHUB account. Create one here: https://nanohub.org/register/
Request membership in the MEST group in nanoHUB at https://nanohub.org/groups/mest

Previous
Previous

Introduction to Large Language Models (LLMs)