A Smart Privacy-Preserving Learning Method by Fake Gradients to Protect Users Items in Recommender Systems

Algorithm 2

Noisy matrix factorization with fake gradient.

	Input: Redefined iteration number k, learning rate , probability p for Random Response and Standard deviation of Gaussian distribution .
	Output: Item profile matrix V
	For all items j, use the probability p Random Response method to estimate the ratio of the users with as .
	Randomly initialize for all i and j.
	fordo
	Initialize , for all j = 1, 2, …, n in central server.
	fordo
	On user i: sample B items uniformly from{}
	fordo
	ifthen
	Draw
	end
	else
	ifthen
end
else
	end
	Draw
	end
	end
	.
	end
	fordo
	ifthen
	end
	else
	end
	end
	fordo
	Update on the local device by gradient descent.
	end
	end