Closed ThomasColthurst closed 14 hours ago
My thought is that even if the data is stored in file as "0", "1", "2", etc., this distribution class plus a simple string emission class would do a good job of modelling typos.
My thought is that even if the data is stored in file as "0", "1", "2", etc., this distribution class plus a simple string emission class would do a good job of modelling typos.