Closed divyajyotiuk closed 3 years ago
@divyajyotiuk Thanks for the contribution ! Its nice that you coded kNN from scratch. Two suggestions/comments:
Could you put the data in the exercise somehow? I think the best way would be to delete the solutions from the notebook, but kept the beginning part where the data is. You could keep/write some comments to help learners code the algorithm, if you want to.
Do you happen to know how to import knn from sklearn ? I think you could also (maybe at the end of the solution) illustrate how to use the kNN package from sklearn.
If I am not wrong, do you want me to [ ] delete the solutions folder altogether ? [ ] Move the sample solution to data folder ? where do I write some comments to help learners code the algorithm - in the sample solution file or make a different file ?
I'll add the second point as you mentioned.
@divyajyotiuk what I meant was to have another notebook in the exercise folder, which is a copy of your solution notebook, but contain the data and/or part of the rest of the codes (but with comments/suggestions on the codes).
I have added starter exercise and also shown an example of using sklearn's knn module.
Thanks @divyajyotiuk, I'll review it later this week.
Thanks @divyajyotiuk , I'll merge now.
Hey, the file name for scikit learn knn code is wrong.
knn_from_scratch.ipynb
should be replaced with knn_using_sklearn.ipynb
in the solution markdown.
https://github.com/gimseng/99-ML-Learning-Projects/tree/master/010/solution
oops i copied pasted and forgot to change. Thanks @divyajyotiuk !
Reference Issues/PRs
This can be a foundation for the #94 issue.
What does this implement/fix? Explain your changes.
Implementation of kNN from scratch and extendible also tasks provided as exercises.
Any other comments?
Beginner friendly algorithm for kNN provided.