Scribd
Upload
20 views1 page

Notes

Uploaded by

Saptarshi Das
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as TXT, PDF, TXT or read online on Scribd
20 views1 page

Notes

Uploaded by

Saptarshi Das
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as TXT, PDF, TXT or read online on Scribd

- standard scaler uses standardization method to scale

- scaling is important to find the relationship without any much difference

https://youtu.be/wZOFkLlRa8c?si=oyCHFQ1P7BsxUMAm

1. Unbalanced data:-

- Oversamapling (Duplicate minority class)

- Undersampling (Minimize the majority class i.e majority class length = min
class length)

- SMOTE: Synthetic Minority Oversampling Technique:-


Oversampling but without duplicacy it generates new datapoints
Uses interpolation
a. Train a KNN on minority
b. Select examples from the minority class at random
c. Select a neighbour of each example at random (for interpolation)
d. Extract a random number between 0 and 1
e. Calculate the new examples as

original sample - factor*(original sample - neighbour)


f. The final dataset consists of the original dataset + the newly created
examples

You might also like