Predict the happiness

This seems like a great competition to work on to practice the skills that Jeremy taught this week. Figured I would share so everybody can benefit. I don’t really know much about the site besides it isn’t Kaggle, but I have done one other competition on there and it seems like a good site (though not as popular as Kaggle) (Also not as good as Kaggle).


I believe this dataset can be used for the Structured dataset problem. Is that right?

Didn’t look at the description field. Makes sense for an NLP case

My first submission. Hoping to improve, but happy with that either way. :slight_smile:


Was that based on a similar technique to our IMDB notebook?

1 Like

Yeah. Pretty much identical.

1 Like


Am I missing something or doing something seriously wrong??

How did you create the learner…

I used
Language ModelData.from_dataframes(train_df,test_df,val_df.)

but after doing that it says you can’t do

(Traceback)(it’s a very long one like a chain reaction…)

> Operation between Nonetype and int

Is there something else needs to be done…

  • one which I thought as a possibility is to convert reviews to.txt files??(isn’t feasible)