Closed Sum02dean closed 2 years ago
This is an attempt to prevent the stochastic results, potentially due to the random sampling of data especially when using COGS. Generate n-datasets, then average the results over all of the n-splits. This may provide some robustness to the model.
This is an attempt to prevent the stochastic results, potentially due to the random sampling of data especially when using COGS. Generate n-datasets, then average the results over all of the n-splits. This may provide some robustness to the model.