We are pleased to announce that the course, Applied Data Science and Machine Learning, will be offered again in the Oregon Health & Science University (OHSU) Spring Quarter of 2022. The course, which now has the official OHSU course number BMI 527/627, is designed for students in the Health & Clinical Informatics (HCIN) Major of the OHSU Biomedical Informatics Graduate Program.
The course provides an overview of the application of data science, machine learning, and artificial intelligence (AI) in health care settings. Students are introduced to a wide range of machine learning topics, including identifying health care issues that can be addressed with machine learning solutions, machine learning model development and data source identification, machine learning model implementation, critical appraisal of machine learning literature, and ethical considerations for the application of machine learning and AI in health care. Students also identify an issue in health and develop their own machine learning model to address this issue. With some constraints, they are welcome to use their own data set.
The course is particularly aimed at HCIN students who will need to implement and critically evaluate the impact of AI systems in health care. It is designed for those who may not have the math background that is required to develop machine learning applications. The course has a prerequisite of BMI 540, which covers computer science and Python programming.
The topical outline of BMI 527/627 includes:
- Overview of biomedical data science
- Overview of biostatistics, machine learning and artificial intelligence
- Critical assessment of machine learning literature – both development and implementation
- Introduction to data sources and programming languages
- Data preparation
- Data exploration
- Using automated model development software (e.g., RapidMiner or Orange) to evaluate strengths and weaknesses of different machine learning algorithms (e.g., kNN, logistic regression, decision trees, random forest, support vector machines and simple neural networks)
- Model implementation
- Ethical considerations
Reading assignments for the course come from two volumes:
- Hoyt R and Muenchen R, Introduction to Biomedical Data Science, Lulu.com, 2019
- Hoyt R and Muenchen R, Data Preparation and Exploration: Applied to Healthcare, Lulu.com, 2019
The course includes content providing programming and modeling skills development. There are weekly assignments in Python or use of automated model development software, such as RapidMiner or Orange. The course aims to give students the necessary skill development for application to each phase of the class project. The figure below depicts the model development process followed by the course.