andrew12·15 年前·議論I guess they meant. They took a random data set and could classify 62% of the names in this data set. The rest, owing to insufficient size of the training data, could not be classified as either.