1 About the course

Machine learning describes a series of data-driven algorithmic approaches that simulate the “learning without being explicitly programmed” paradigm. These methods are particularly useful when limited information is available about the structure or properties of a dataset; also, real-world data rarely follows a well-defined mathematical distribution (due to technical variation in measurements, noise, etc) – assumption-free models offer flexibility for this type of input with the side effects of underlying characteristics of the dataset (e.g. through feature selection). The term “Machine Learning” encompasses a broad range of approaches in data analysis with wide applicability across biological sciences. Lectures will introduce commonly used approaches, provide insight into their theoretical underpinnings and illustrate their applicability and limitations through examples and exercises. During the practical sessions students will apply the algorithms to real biological data-sets using the R language and RStudio environment. All code utilised during the course will be available to participants.

1.1 Events Page

Course info can be found here

1.2 Prerequisites

Participants should be experienced in programming in R as the course will build on this. We recommend the Introduction to R for biologists course as a first course to start programming in R. If you are not able to attend an introductory course, please work through the R material as a minimum.

1.3 Schedule

Time Data Module
09:30 – 10:00 17/12/21 Introduction lecture to regression
10:00 – 11:00 17/12/21 Linear regression / linear models
10:00 – 11:00 17/12/21 Coffee break
11:15 – 12:00 17/12/21 Logistic regression
12:00 – 1:00 17/12/21 Lunch break
1:00 – 1:45 17/12/21 Introduction lecture to neural networks
1:45 – 2:45 17/12/21 Artificial Neural Networks
2:45 – 3:00 17/12/21 Coffee break
3:00 – 4:30 17/12/21 Convolutional neural nets and beyond
4:30 – 5:00 17/12/21 Summary lecture and questions

1.4 Github

A link to the GitHub code for day 3 of this course can be found here.

1.5 Google docs interactive Q&A

An interactive Google docs link for Q&A can be found here here

1.6 License

GPL-3

1.7 Contact

If you have any comments, questions or suggestions about the material for day 3, please contact Chris Penfold.

1.8 Colophon

This book was produced using the bookdown package (Xie 2017), which was built on top of R Markdown and knitr (Xie 2015).

References

Xie, Yihui. 2015. Dynamic Documents with R and Knitr. 2nd ed. Boca Raton, Florida: Chapman; Hall/CRC. http://yihui.name/knitr/.

Xie, Yihui. 2017. Bookdown: Authoring Books and Technical Documents with R Markdown. https://github.com/rstudio/bookdown.