[Google_Bootcamp_Day2]

Logistic regression recap

Logistic regression derivatives

1 single step of Gradient descent on m examples

• 2 for loops (1 for m training examples and 1 or number of features) are inefficient
• Needs to be vectorized implementation

Vectorization

• In neural network programming, whenever possible, avoid explicit for-loops

Logistic regression derivatives

Vectorizing Logistic regression

• Before vectorization (with for loop)

• After vectorization (without for loop)

Vectorizing Logistic regression’s gradient computation

Implementing Logistic Regression

[Source] https://www.coursera.org/learn/neural-networks-deep-learning