@yenisel Probably depends on data and model, but you might extrapolate performance from smaller sample sizes, e.g. https://t.co/olyCPOJ1Ht
RT @triadsou: Predicting sample size required for classification performance. Rosa L Figueroa, Qing Zeng-Treitler, Sasikiran Kandula & Long…
Predicting sample size required for classification performance. Rosa L Figueroa, Qing Zeng-Treitler, Sasikiran Kandula & Long H Ngo. BMC Medical Informatics and Decision Making 2012; 12: 8. https://t.co/yo9jimWw8J
RT @DrJohnKang: Thanks and nice ref! Learning curve for several ML classifiers (at least the non-deep) obeys inverse power law. Important t…
Thanks and nice ref! Learning curve for several ML classifiers (at least the non-deep) obeys inverse power law. Important to cite this and Mukherjee papers when justifying sample size for ML models.
RT @kdpsinghlab: @DrJohnKang I agree that if you have access to data, then a learning curve can be instructive. Found this reference useful…
@DrJohnKang I agree that if you have access to data, then a learning curve can be instructive. Found this reference useful (for classification problems). https://t.co/KeqAm29ZJj
Predicting sample size required for classification performance: Rosa L Figueroa, Qing Zeng-Treitler,... http://t.co/SbaQNar30E #HealthIT
Predicting sample size required for classification performance http://t.co/0LQfS1OG @BMC_Series #MedicalInformatics #machinelearning