Pistoia Alliance April Virtual Conference: Federated Machine Learning On Heterogenous Drug Discovery Data

This on-demand recording is part of the Pistoia Alliance Conference: Collaborative R&D in Action, April 20-23, 2021. For more information about related events, please visit our online calendar.

Summary

Data relevant for building robust drug discovery models is distributed within and across companies. Federated Learning is a machine learning technique to train an algorithm across multiple decentralized datasets which increasingly finds adoption.

In this session, we will take a look at different techniques to solve one of the biggest bottlenecks of federated data set-ups: heterogenous datasets.

We will discuss how data can be normalized and harmonized across different parties before training and how to train machine learning models on data of different modalities.