Hello, you mentioned in the text that when predicting the band gap, only data greater than 0 is used. So, in the context of machine learning regression, if the target value to be predicted (in this case, the band gap) has half of the original data as zero, what methods would you suggest for handling this situation?
Hello, you mentioned in the text that when predicting the band gap, only data greater than 0 is used. So, in the context of machine learning regression, if the target value to be predicted (in this case, the band gap) has half of the original data as zero, what methods would you suggest for handling this situation?