Ethan Nam -Stats and Learning (1)
.pdf
School
University of Illinois, Urbana Champaign**We aren't endorsed by this school
Course
BADM 322
Subject
Economics
Date
Dec 18, 2024
Pages
2
Uploaded by SargentMoonSeaUrchin32
In terms of the first monkey learning data, one thing I noticed was the lower accuracyrate and F1 score percentage. Even though there was lower text within the sample data, Ibelieve the accuracy rates and the F1 score percentage were lower because the programwasn’t able to analyze the ratings as well as the other models. I believe if the data was cleanerand we ran more tests, we would have a higher accuracy percentage score. In terms of thesecond model, I believe the accuracy score and the F1 percentage were very high due to thefact that the data was cleaner. Although there were more texts, I believe the high accuracyscore represents how good the data was, therefore, making it easier to read. Finally, with thelast model, I believe the accuracy score was high due to the smaller batch size. As a result, itmade it easier for Monkey Learn to analyze the data.