Transparency and reproducibility in artificial intelligence
Ontario Institute for Cancer Research · University Health Network · +35 more institutions
Abstract
Breakthroughs in artificial intelligence (AI) hold enormous potential as it can automate complex tasks and go even beyond human performance. In their study, McKinney et al. showed the high potential of AI for breast cancer screening. However, the lack of methods’ details and algorithm code undermines its scientific value. Here, we identify obstacles hindering transparent and reproducible AI research as faced by McKinney et al., and provide solutions to these obstacles with implications for the broader field.
Citation impact
- FWCI
- 16.99
- Percentile
- 100%
- References
- 18
Authors
32- BHBenjamin Haibe‐KainsCorresponding
Ontario Institute for Cancer Research, University Health Network, University of Toronto, Princess Margaret Cancer Centre, Vector Institute
- GAGeorge Alexandru Adam
University of Toronto, Vector Institute
- AHAhmed Hosny
Brigham and Women's Hospital, Harvard University, Dana-Farber Cancer Institute, Dana-Farber Brigham Cancer Center
- FKFarnoosh Khodakarami
University Health Network, University of Toronto, Princess Margaret Cancer Centre
- MAMassive Analysis Quality Control (MAQC) Society Board of Directors
National Center for Toxicological Research, United States Food and Drug Administration
Topics & keywords
- Transparency (behavior)
- Computer science
- Field (mathematics)
- Artificial intelligence
- Data science
- Computer security
- Mathematics
- Gender equality