AI with Matlab

AI plays an important role in automating workflows in value chains and industrial processes. Recently, AlphaFold had a huge impact on structural biology and will change this field forever. This course makes you familiar with the basic mathematical principles of AI.

We start with building a single neuron, expand it to a layer and add all the optimization we need for a fully functional network. We will train our own network, run pretrained networks provided by Matlab, work with tools like the image labeler and apply it to real Cryo-EM (Beckmann lab, see also here) and optical image (Gompel lab) data sets.

Topics:

• creating a neuron
• building a layer of neurons
• activation functions
• gradient descent
• backpropagation
• the training process
• CNNs in Matlab
• pretrained NN in Matlab
• Segmentation in Matlab