MLx Fundamentals:

Based on the success of previous years' program, and in order to provide all participants with the necessary background -- particularly for those who are new to the theory and fundamentals of modern ML --  during this module, we aim to provide everyone with training in the following topics:

 

• Linear Algebra and Mathematics of machine learning

• Optimisation

• Fundamentals of statistical / probabilistic ML

• Fundamentals of representation / deep learning

• and more

​

MLx Cases:

The aim of ML x Cases track is to provide you with a training on real-world issues and processes related to ML development/implementation process. This will range from efficient and repeatable approaches to data collection, enrichment and cleaning, and labelling, to transfer learning use cases of pre-trained SOTA models and their fine-tuning to achieve good performance on a domain-specific task. We will run ~5 different cases, led by experienced ML / data scientists, supported by TAs to help make the sessions interactive.
At the end of the ML x Cases, participants will learn useful concepts on:

​

• Frame a problem as an ML problem

• Leveraging appropriate toolboxes

• Knowing which approach typically works best depending on the types of use cases

• Defining what performance metrics to choose

• Experimental setups for a performant model, while tracking and documenting experiments with MLFlow

• Forming a naive baseline to more sophisticated experiments

• Interpreting model results (e.g., under/overfitting and ways to remediate it).

• Feedback loops and allowing the system to collect information from user inputs.

• and more

​