Introduction to machine learning
Understand the power, and pitfalls, of machine learning for data-driven decision making.
Machine learning has immense potential across a broad range of applications: from computer vision to prediction modelling and natural language processing. But in recent years, the hype surrounding machine learning has led to inflated expectations in some settings. And while advances in computing and code-sharing have significantly lowered the barriers to applying machine learning techniques, they have also caused an increase in the number of users applying machine learning suboptimally, or even inappropriately.
Researchers, clinicians and industry executives all stand to benefit from from a basic understanding of machine learning processes, including when and how they can be deployed for maximum effect, and when other techniques would be more appropriate.
In this one-day remote teleconference course, we cover the fundamental principles of machine learning including:
What is meant by machine learning
An overview of common machine learning models - how they work and where they can be used most effectively
Techniques for training machine learning algorithms and a typical development pipeline, from data collection to external validation and post-market monitoring
How machine learning algorithms are evaluated and why certain performance metrics can be misleading
Opening black boxes with explainable AI
Current controversies in machine learning and prediction modelling more broadly
Guidelines for performing, reporting and interpreting machine learning studies
The course focuses on fundamental concepts and requires no background in statistics or programming. By the end, you will have an understanding of how common machine learning algorithms work, where they should and should not be applied, be able to critically appraise machine learning studies, and feel ready to take the next step towards practical machine learning in Python or R.
