Machine Learning for Sciences (MLS)

Welcome to the interactive laboratory for the Machine Learning in Sciences (MLS) course at UCPH. These tools are designed to provide intuition about fundamental ML concepts that will be covered in the course.

• Supervised Learning Linear Classifier Try out the influence of different parameters of the linear classifier to obtain good performance.
• Supervised Learning k-Nearest Neighbour Interactively see the influence of changing k on the classification decision boundary.
• Supervised Learning Decision Trees Partition feature spaces to engage with decision trees.
• Unsupervised Learning Representation Learning Learn a representation space for high-dimensional objects like images.
• Basics Curse of Dimensionality See how dimensionality impacts distances.
• Basics Gradient Descent Try out how learning rate and initialisation impact gradient descent algorithm.
• Basics Backpropagation See how gradients are computed using chain rule in computational graphs.
• Unsupervised Learning Clustering Try out k-means clustering.
• Unsupervised Learning Principal Component Analysis Try out PCA and explore explained variance.
• Basics Activation Functions See different activation functions and their derivates.
• Basics Convolution Operation in CNNs Try out how CNNs learn and extract features.
• Basics Message Passing in Graph Neural Nets Try out message passing in a simple graph.
• Basics Self-Attention Coming Soon