A Bitwise Design and Implementation for Privacy-Preserving Data Mining: From Atomic Operations to Advanced Algorithms

Algorithm 3

The algorithm of linear regression.

	Input: data , , learning rate α, number of iteration
	Output: Parameter θ
(1)	Initialize parameter θ to FX
(2)	Gradient descent part 1: calculate partial derivative of cost function
(3)	Gradient descent part 2: multiply α with the value of part 1
(4)	Gradient descent part 3: update θ until iteration times
(5)	return each of θ’s