Data Science Machine-Learning Science Scientific Libraries (Numpy/Pandas/SciKit/...)See in schedule
AutoML makes machine learning accessible for everyone!
Machine Learning is hard since developing well-performing machine learning pipelines requires a lot of expertise, time and manual tuning. AutoML automates this development process by using latest optimization methods to efficiently search for well performing solutions.
In this talk, we will cover how to move from manually constructing and tuning machine learning pipelines to using efficient hyperparameter optimization algorithms and full AutoML using the popular open-source Auto-sklearn library. Auto-sklearn is a drop-in replacement for any scikit-learn estimator and is developed by the ML Lab of the University of Freiburg.
More specifically, you’ll learn the following:
* What are algorithm portfolios and why are they useful?
* What is Bayesian optimization and for what can you use it?
* How to use Auto-sklearn instead of your scikit-learn estimator.
This talk assumes basic understanding of machine learning and statistics. The main target audience are data scientists and domain experts using machine learning. The talk will be designed such that anyone with a basic understanding of machine learning pipelines in scikit-learn and the Python language would be able to understand the concepts and to use our tool.
Type: Talk (45 mins); Python level: Beginner; Domain level: Intermediate
Matthias Feurer is a doctoral candidate at the Machine Learning Lab at the University of Freiburg, Germany. His research focuses on automated machine learning, hyperparameter optimization and meta-learning. He is actively involved in developing open source software for AutoML and is the maintainer and co-creator of Auto-sklearn and OpenML-Python. Matthias is a founding member of the Open Machine Learning Foundation, gave AutoML tutorials at the GCPR and ECMLPKDD summer school and co-organized the AutoML workshop in 2019 and 2020. Furthermore, he was part of the winning team of the 1st&2nd AutoML challenges and the BBO challenge@NeurIPS 2020.
Katharina Eggensperger is a PhD student at the Machine Learning Lab at the University of Freiburg, Germany. Her research focuses on empirical performance modeling, automated machine learning and hyperparameter optimization. She has been an invited speaker at the BayesOpt workshop at NeurIPS 2016 and co-organized the AutoML workshop in 2019, 2020 and 2021.