We use cookies to ensure you have the best browsing experience on our website. Please read our cookie policy for more information about how we use cookies.
Stitch the Torn Wiki
Stitch the Torn Wiki
Sort by
recency
|
6 Discussions
|
Please Login in order to post a comment
I won't post the code here, but here's how I solved it:
Got 100% with this approach
For help, you can follow up the link:
https://github.com/nishant-sethi/HackerRank/
Hmm. With every optimization I can think of and a conflict-resolution scheme between the matches I'm still stuck at 96.67%. Really quite discouraging. Why's the solved score set so high?
Edit: got the 100 with a small tweak, but I'd still suggest adding a little leeway in the score.
Hi,
I have tried cleaning the data using NLTK, tfidf (with increasing grams), removed very frequent words and SVD. However, I'm still stuck at 95% as it seems classes 9 and 16 are extremely similar. Could someone recommend a better approach?
@PRASHANTB1984 It is written that
nltk
library is supported, but I can not use a lot of the functionality from it. Like for examplefrom nltk.corpus import stopwords
will end up with:Resource u'corpora/stopwords' not found. Please use the
NLTKDownloader to obtain the resource: >>> nltk.download()
It would be nice if someone from hackerrank would be able to resolve this problem.