@misc{kapoor_irreproducible_2021,
	title = {({Ir}){Reproducible} {Machine} {Learning}: {A} {Case} {Study}},
	abstract = {The use of Machine Learning (ML) methods for prediction and forecasting has become widespread across the quantitative sciences. However, there are many known methodological pitfalls in ML-based research. As a case study of these pitfalls, we examine the subfield of civil war onset prediction in Political Science. Our main finding is that several recent studies published in top Political Science journals claiming superior performance of ML models over Logistic Regression models fail to reproduce. Our results provide two reasons to be skeptical of the use of ML methods in this research area, by both questioning their usefulness and highlighting the pitfalls of applying them correctly. Results identifying errors in studies that use ML methods have appeared in at least seven quantitative science fields. However, we go farther than most previous research to investigate whether the claims made in the reviewed studies survive once the errors are corrected. We argue that there is a reproducibility crisis brewing in research fields that use ML methods and discuss a few systemic interventions that could help resolve it.},
	language = {en},
	author = {Kapoor, Sayash and Narayanan, Arvind},
	pages = {6},
	year = {2021},
	howpublished = {\url{https://reproducible.cs.princeton.edu/}},
	url = {https://reproducible.cs.princeton.edu/},
	urldate = {2021-07-28}
}