BME 6790

Machine Learning and Neural Network Design for Biomedical Applications

Course Description

Course information provided by the 2026-2027 Catalog.

This one semester course will be focused on exposing students to basic strategies in machine learning using neural networks within the context of biomedical engineering. This includes early uses (classical fitting), and basic to concepts such as loss functions, models, backpropagation and training, as well as current layer (dense, convolutional networks and x-formers), and model architectures (e.g. autoencoders, U-Nets, adversarial networks, and large language models) and how these are applied towards current biomedical engineering tasks (medical image recognition, bioinformatics, etc.). This will be geared towards students who are interested in learning to design, code and understand common neural network strategies. Course materials will be primarily implemented in Python, using common packages, such as NumPy, SciPy, Pandas, and TensorFlow, in addition to open-source databases. Students are expected to have a basic familiarity with python programming and some experience with applying statistical methods.

Prerequisites CS 1110 or equivalent, ENGRD 2700, BTRY 3010, BTRY 3020, or ILRST 2110, or equivalent.

Enrollment Priority Recommended prerequisite: MATH 2940.

Last 4 Terms Offered (None)

Learning Outcomes

• Demonstrate the ability to distinguish between different types of modern neural network architectures and the kinds of physical problems they can be applied to.|Explain the fundamental strengths and weaknesses of current neural network models, and how they contribute to their function and limitations (e.g. For instance, how are LLM trained, how does this limit them or contribute to the type of information they can provide).|Demonstrate the ability to implement and train a basic neural network model as applied towards a biomedical application

View Enrollment Information