Learning Task with Future Leaked Information
Jump to navigation
Jump to search
An Learning Task with Future Leaked Information is a learning task where the modeler has visibility to unknowable information.
- See: Lookahead Bias, Learning Dataset.
References
2017
- https://www.kaggle.com/alexisbcook/data-leakage
- QUOTE: ... But the model will be very inaccurate when subsequently deployed in the real world, because even patients who will get pneumonia won't have received antibiotics yet when we need to make predictions about their future health.
To prevent this type of data leakage, any variable updated (or created) after the target value is realized should be excluded. ...
- QUOTE: ... But the model will be very inaccurate when subsequently deployed in the real world, because even patients who will get pneumonia won't have received antibiotics yet when we need to make predictions about their future health.
2011
- (Kaufman et al., 2011) ⇒ Shachar Kaufman, Saharon Rosset, and Claudia Perlich. (2011). “Leakage in Data Mining: Formulation, Detection, and Avoidance.” In: Proceedings of the 17th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD-2011) Journal. ISBN:978-1-4503-0813-7 doi:10.1145/2020408.2020496
2010
- Rosset, S., Perlich, C., Swirszcz, G., Liu, Y., and Prem, M. 2010. Medical data mining: lessons from winning two com-petitions. Data Mining and Knowledge Discovery. 20(3) 439-468.