HBA1C and CREATININE are the two criteria that are value-based and a majority classifier (#3) doesn't work.
It could be that machine learning approaches (#10 or #4) also doesn't perform well. In that case, we might need to parse the text to extract the corresponding values.
HBA1C and CREATININE are the two criteria that are value-based and a majority classifier (#3) doesn't work.
It could be that machine learning approaches (#10 or #4) also doesn't perform well. In that case, we might need to parse the text to extract the corresponding values.