Chain Rule Through One Sigmoid Neuron
~12 mincode completion
Implement neuron_gradients(w, b, x, y) returning the array [da/dw, dL/dw].
Where this shows up
~12 min
••••••••
4 employers weight this skill
4 quant funds. Top match scores 93.
Python
import numpy as np
def neuron_gradients(w, b, x, y):
"""
Derivatives of one sigmoid neuron's output and squared error with respect to w.
Args:
w: weight
b: bias
x: input
y: true answer
Returns:
array [da/dw, dL/dw]
"""
# YOUR CODE HERE
pass