Reproducibility Requires Consolidated Artifacts
This addresses the reproducibility problem for the machine learning research community, but it is incremental as it builds on existing analyses and proposes a known solution.
The paper tackles the reproducibility crisis in machine learning by analyzing 142 replication studies and 204 code repositories, finding that missing experiment details like hyperparameters cause failures, and concludes that consolidated artifacts with a unified framework can improve reproducibility.
Machine learning is facing a 'reproducibility crisis' where a significant number of works report failures when attempting to reproduce previously published results. We evaluate the sources of reproducibility failures using a meta-analysis of 142 replication studies from ReScience C and 204 code repositories. We find that missing experiment details such as hyperparameters are potential causes of unreproducibility. We experimentally show the bias of different hyperparameter selection strategies and conclude that consolidated artifacts with a unified framework can help support reproducibility.