SMOTE: Finding Nearest Neighbors
Beginner
The problem, from first principles
SMOTE needs the nearest minority neighbors of each sample. The paper uses Euclidean distance, and the neighbors are chosen from the minority class only.
From theory to code
Implement nearest_neighbors(X, x, k), returning the indices of the k closest rows, nearest first.
Constraints
- Distances are Euclidean.
Hints
Hint 1
Compute the distance from x to every row, then sort the indices by distance and keep the first k.
Keywords
research-papersclassical-mlimbalancesmote
Python
Run your code to see results
No results yet
Run your code to check it against the examples, or submit to run every test.