Abstract
Automated machine learning (AutoML) supports the algorithmic construction and data-specific customization of machine learning pipelines, including the selection, combination, and parametrization of machine learning algorithms as main constituents. Generally speaking, AutoML approaches comprise two major components: a search space model and an optimizer for traversing the space. Recent approaches have shown impressive results in the realm of supervised learning, most notably (single-label) classification (SLC). Moreover, first attempts at extending these approaches towards multi-label classification (MLC) have been made. While the space of candidate pipelines is already huge in SLC, the complexity of the search space is raised to an even higher power in MLC. One may wonder, therefore, whether and to what extent optimizers established for SLC can scale to this increased complexity, and how they compare to each other. This paper makes the following contributions: First, we survey existing approaches to AutoML for MLC. Second, we augment these approaches with optimizers not previously tried for MLC. Third, we propose a benchmarking framework that supports a fair and systematic comparison. Fourth, we conduct an extensive experimental study, evaluating the methods on a suite of MLC problems. We find a grammar-based best-first search to compare favorably to other optimizers.
Highlights
AUTOMATED machine learning (AutoML) is commonly understood as the task of automating the process of engineering a “machine learning pipeline” tailored to a problem at hand, that is, to a dataset on which a model ought to be induced
We considered existing optimization approaches for automating multi-label classification and, transferred other AutoML approaches commonly used for singlelabel classification to the problem domain of MLC
Our extensive study revealed that a reduction of the AutoML problem to hyper-parameter optimization does not scale well to the problem domain of MLC out of the box
Summary
AUTOMATED machine learning (AutoML) is commonly understood as the task of automating the process of engineering a “machine learning pipeline” tailored to a problem at hand, that is, to a dataset on which a (predictive) model ought to be induced. This includes the selection, combination, and parameterization of machine learning (ML) algorithms as basic constituents of the pipeline, which is the main output produced by an AutoML tool, and which can be used to train a concrete model on the dataset. Since an AutoML tool is a complex system consisting of several components, most importantly a search space model and an optimization method for traversing this space, one typically faces a credit assignment
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: IEEE transactions on pattern analysis and machine intelligence
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
