#24 Machine Learning Specialization [Course 1, Week 2, Lesson 1]  Summary and Q&A
TL;DR
Implementing gradient descent for multiple linear regression using vectorization to efficiently optimize the cost function.
Key Insights
 💄 Multiple linear regression can be represented more succinctly using vector notation, making it easier to implement and optimize.
 🪡 Gradient descent needs to be modified when there are multiple features in the regression model.
 🔙 The normal equation method is an alternate algorithm for finding W and B in linear regression but is not suitable for other learning algorithms.
 🐢 The normal equation method can be slow for large numbers of features.
 😒 Most machine learning libraries use gradient descent for linear regression, but some may implement the normal equation in the backend.
Transcript
so you've learned about gradient descents about multiple linear regression and also vectorization let's put it all together to implement gradient descent for multiple linear regression with factorization this would be cool let's quickly review what multiple linear regression look like using our previous notation let's see how you can write it more ... Read More
Questions & Answers
Q: How is multiple linear regression represented using vector notation?
In vector notation, the parameters of multiple linear regression are collected into a vector W and a number B. The model can be written as F sub w b of x = w dot product with x + b.
Q: How is the cost function defined for multiple linear regression?
The cost function J is defined as a function of the parameter vector W and the number B. It takes input as a vector W and the number B and returns a number.
Q: How does gradient descent update the parameters in multiple linear regression?
In gradient descent, each parameter WJ is updated using the formula WJ = WJ  Alpha * derivative of the cost function J with respect to W. The update rule for parameter B is also separate.
Q: What is the normal equation method for finding W and B in linear regression?
The normal equation method is an alternative way to find W and B that does not require iterative gradient descent. It uses advanced linear algebra to solve for W and B directly.
Summary & Key Takeaways

Multiple linear regression can be represented using vector notation, where the parameters are collected into a vector W and a number B.

The cost function is written as a function of the parameter vector W and the number B.

The update rule for gradient descent is different when there are multiple features compared to just one feature.