⚠️ A valid COVID certificate must be presented on site to enter the event. ⚠️

Datasets of interest in many application domains (e.g. healthcare, finance data, 
manufacturing) contain sensitive or private information and cannot easily be shared. 
Additionally, such data frequently belongs to multiple distinct parties and 
combining it in one location would expose a lucrative target to hackers. 
Therefore, it is desirable to make use of such data without a need 
to disclose it or store it in a central location. 

Unfortunately, traditional methods to train predictive models expect data to 
be fully accessible and centralized on a single server. 
Research work therefore has to rely on small or artificial datasets that 
can safely be centralized. As a result, findings frequently do not generalize well 
to real-world datapoints and progress is hampered. 

Federated Learning (FL) is a recently introduced paradigm that addresses this limitation by training models on 
decentralized datasets without requiring centralized data access. 
This approach allows multiple distinct parties to collaboratively train predictive 
models without a need to directly share sensitive data. Instead of combining 
datasets, FL trains a model in multiple iterations on data subsets stored 
in different locations. 
In every iteration, every party owning a data subset downloads a copy of the current model weights. 
An updated model is computed for each data subset in a local training step. 
Per-party model updates are then aggregated (a step that can be centralized, as it does not require 
data access) resulting in a single overarching FL update step. 

As an introduction to the workshop, we will introduce the basic concepts underlying 
FL and discuss a few of the key related topics (e.g. Differencial Privacy, 
Model Encryption). Our focus however, will be on gaining hands-on experience. 
We will implement a simple Federated Learning system using tensorflow (tensorflow/federated) and pytorch (PySyft). 
We will give a quick introduction to all needed libraries and tools at the start.

Teach your models how to use decentralized, sensitive data without a need to have any direct access to it. 

Hands-on is key: learn how to train models using Federated Learning in tensorflow or pytorch.

For technical preparations please follow instructions of  the "Preparation" section on the following page.

Participation with a laptop is recommended as a large part of this workshop will be hands-on.