Online: Ensuring More Accurate, Generalisable, and Interpretable Machine Learning Models for Bioinformatics
This course is addressed to life scientists, bioinformaticians, and computational biologists who would like to learn more about general best practices in Machine Learning and get more out of their Machine Learning models: more precise hyper-parameters, more generalizable models, and more interpretable models.
Application deadline: 29 September 2024
Date: 15 October 2024
General information
- Location: Online
- Application deadline: 29 September 2024
- Date: 15 October 2024
- Level: Intermediate
- Length: 8h
- Fees:
- Academic: 100 CHF
- For-profit: 500 CHF
Description
Machine Learning has become an essential tool in Life Science, letting scientists explore and learn from large and complex biological datasets. To collectively unravel the puzzle of life, we must ensure that machine learning models make the most of the available data and that they are correctly generalizable, robust, and interpretable to provide trustworthy and actionable insights. This advanced course is designed for scientists who already have a foundational understanding of machine learning and seek to enhance their core skills in this domain.
This course focuses on best practices and advanced techniques in Machine Learning, aiming to provide you with the tools needed to develop more accurate, generalizable, and transparent models.
Prerequisites
Please check the numerous prerequisites here.
Learning outcomes
At the end of this course, you will be able to:
- Use the hyperopt library to efficiently explore your hyper-parameter space with Bayesian Optimization and tune your models.
- Evaluate the generalizability of your generated models using best practices such as nested cross-validation.
- Explain the role of each feature in your model’s prediction, even for so-called “black-box” models
- Examine the results of your models and assess their quality.