Konferenz: Quantitative Network Science


Cascade processes in machine learning

{{_Ltalk:R}} Dr. Rebekka Burkholz
Datum: 26.11.20   Zeit: 14.00 - 14.45   Raum:

Cascade processes have been proven useful in modeling such diverse phenomena as epidemic spreading, signaling in biological networks, information propagation in social media, financial systemic risk, and the reorganization of international trade networks. In this talk, we will focus on two specific model classes and use them as means to an end to improve and develop machine learning algorithms. We will leverage the fact that load redistribution models correspond one-to-one to the evaluation of deep neural networks. Based on this insight, we will derive successful initialization strategies for deep learning and speed up model training based on analytic insights for random graph ensembles. We will further discuss how we can utilize these improvements for the inference of gene regulatory networks.