Flashcards
ex2 ml lectNote
Cards in this deck (19)
- 1multivariable hypothesis function
- 2multiple feature hypothesis function
- 3multivariable linear regression
- 4training example X row-wise
- 5hypothesis calculation = column vector of size mx1
- 6cost function
- 7cost function - vectorized version
- 8gradient descent for multiple variables
- 9in other words, gradient descent
- 10matrix notation gd rule can be expressed
- 11The j-th component of the gradient is the summation of the product of two terms:
- 12product of two terms can be expressed as the product of two vectors
- 13matrix notation of GD rule finally i is:
- 14feature normalizationspeed up gradient descent by having
each of the input values in roughly
the same range.
because theta will descend quickly on
small ranges and slowly on large ranges - 15feature normalization ideally"
- 16goal of feature normalizationget all input vars into roughly one of
these ranges - 17two techniques to help with arefeature scaling : involves dividing the input variables by the range of the input var, resulting in new range of just 1
mean normalization:subtracting the average value for an input variable from the values for that input var, resulting in a new average value for the input var of just 0 - 18implement both feature scaling x mean normalizationwhere muy_i is the avgerage of all the values for feature(i) and s_i is the range of values(max - min), or s_i is standard deviation
- 19example x_i
Information
- Description
- machineLearning Flashcards on ex2 ml lectNote, created by Ax H on 05/03/2019.
- Created by
- Ax H
- Views
- 0
- Rating