Meta-Learning and AutoML: Introduction to "Learning to learn"

The speaker will talk about learning to learn problem in machine learning. Different ML learning to learn approaches will be introduced such as Transfer Learning, Active Learning, Meta-Learning, etc. Specifically we will talk about meta-learning : what it is and a survey about it. Then we shall talk about AutoML (practical side of learning to learn) and end with ongoing research areas/problems.

Speaker Bio:
Mikhail Mekhedkin-Meskhi
Currently he is working as a Data Scientist at PDR Corp. He is starting his PhD in CS at UH this upcoming August under Dr. Ricardo Vilalta. His main research focus is meta-learning, Topological Data Analysis and AutoML. He also contributes to some open source projects in ML such as OpenML.