Solving ML with ML: Effectiveness of the Metropolis-Hastings algorithm for synthesizing Machine Learning Pipelines

Solving ML with ML: Effectiveness of the Metropolis-Hastings algorithm for synthesizing Machine Learning Pipelines

Sheremet, Denys (TU Delft Electrical Engineering, Mathematics and Computer Science)

Dumančić, S. (mentor)
Hinnerichs, T.R. (graduation committee)
Tax, D.M.J. (graduation committee)

Delft University of Technology

Computer Science and Engineering

CSE3000 Research Project

2023-06-25

In AutoML, the search space of possible pipelines is often large and multidimensional. This makes it very important to use an efficient search algorithm. We measure the effectiveness of the Metropolis-Hastings algorithm (M-H) in a pipeline synthesis framework, when the search space is described by a context-free grammar. We also compare the performance of the M-H algorithm to other search algorithms. While AutoML frameworks use many different search algorithms, and comparisons between AutoML frameworks exist, this is the first paper that compares the performance of different search algorithms in the context of pipeline synthesis under equal conditions. We found that M-H is slightly outperformed by BFS2, the simplest possible search algorithm. We conclude that the datasets we use for evaluating the algorithms are too simple to meaningfully compare the performance of different search algorithms. We also conclude that for simple datasets, simple search algorithms work best.

AutoML
Pipeline Synthesis
Machine Learning
Automated Machine Learning
Program Synthesis

http://resolver.tudelft.nl/uuid:1d0c2a54-6e34-4727-8858-7c6a78968d22

Student theses

bachelor thesis

© 2023 Denys Sheremet