Closed stevehadd closed 4 years ago
This has now been demomstrated in the ML tree cross-validation notebook and should be applied to all other notebooks, including plots of what portion of profiles require the imeta fallback option.
This has now been demonstrated in notebooks and implemented in the experiment framework.
In some cases, the trained ML classifier is not able to produce a classification for a profile. In these cases, we should back on the iMeta value for that profile. This should be reflected in the classification quality flag, using 0 for iMeta and 1 for ML classifier.