5.5 Conclusion

If you found this short walkthrough interesting, you’ll want to check out the full notebooks on the page. They dive deep into using different classifiers, and different ways of counting words.

In the end, though, this airbags research is just a toy example to prove that throwing data science at a problem doesn’t automatically solve it. If we labeled more cases our classifier would probably be better, yes, but in all honestly we should just search for “airbag” and manually read the cases. It’ll take a little more time, but we’ll be able to have much more faith in our end result.