Coding With Data for Biologists (including examples from AI)Info Location Attendee Categories Contact More Info Event Information
DescriptionThis Matlab course assumes no prior coding experience and is aimed at people interested in seeing how ideas from data analysis can be applied to questions in biology and medicine. It is well suited to PhD students and post-doctoral researchers in particular. In person attendance is limited to 40 people.
Event Location
Attendee CategoriesOnline attendance
Additional ItemsContactLisa Butt, More InformationThis Matlab course assumes no prior coding experience and is aimed at people interested in seeing how ideas from data analysis can be applied to questions in biology and medicine. It is well suited to PhD students and post-doctoral researchers in particular. In person attendance is limited to 40 people. The typical tools we’ll use are regression (linear, nonlinear and neural network-based) and clustering (with Gaussian Mixture Models) that we will apply to simple data taken from real-world problems. The latter range from global covid epidemic data that we will study using nonlinear regression to questions about clinical antibiotic resistance data that we will tackle using artificial neural networks. The emphasis of the course will be to make the data and codes as simple as possible, where the problems we tackle are research grade, or close to it. The coding concepts can be taken in a variety of different directions so we’ll discuss ideas from mathematical and statistical modelling from the perspective of, for example, experimental design and P-hacking. We’ll use a few lines of code to help develop intuition on how many replicates you’d need to see convincing, and less convincing, patterns in data. We’ll discuss theoretical models of problems from evolutionary biology and we’ll solve simple differential equations that arise in the pharmacology of antibiotics, often in just a few lines of code. The course is based on a self-contained 250-page PDF of notes that contains all codes, questions with answers and suggestions for small projects. An awful lot can be achieved with just 10 lines of code in Matlab and, in our view, Matlab’s development environment is better suited to the novice coder than, say, Python. For instance, training different neural networks on a dataset can be done with a click of the mouse. By the end of the course, you will be in a position to apply new concepts from data science and Machine Learning that are built into Matlab. Matlab’s language is also Python-like so you’ll also be in a good position to transition to using libraries in Python with Anaconda, which are free. We’ll also show you how to set up Latex, the professional typesetting language that is far, far, superior to anything you can do with Microsoft Word; it’s also completely free.
A laptop with MATLAB already installed is essential | |||||||||||||||
