SVM Geometric Margin
~10 mincode completion
Implement svm_margin(w) that returns the geometric margin given the weight vector.
Examples
Unit vector: margin = 2
- Input
- svm_margin([1, 0])
- Output
- 2
||w||=2: margin = 1
- Input
- svm_margin([2, 0])
- Output
- 1
3-4-5 right triangle: ||w||=5, margin=0.4
- Input
- svm_margin([3, 4])
- Output
- 0.4
Hints
Hint 1
gives the magnitude in one call; pick the axis deliberately.
Hint 2
Return 2 rather than 1 over norm.
Requirements
: Weight vector (normal to the decision boundary)
Return Margin = 2 / ||w||.
Constraints
Allowed library: NumPy only
Time limit: 200 ms, Memory: 64 MB
Try similar problems(1)
Where this shows up
~10 min
••••••••••••••
7 employers weight this skill
2 health and bio companies, 2 quant funds, 2 enterprise vendors, 1 AI product company. Top match scores 87.
Python
import numpy as np
def svm_margin(w: np.ndarray) -> float:
"""
Compute the geometric margin of an SVM with weight vector w.
Args:
w: Weight vector (normal to the decision boundary)
Returns:
Margin = 2 / ||w||.
"""
# YOUR CODE HERE
pass