Consider the Perceptron model,

y =sign(wx(n)), with the learning algorithm. Select all the true statements:

options:

	Perceptron algorithm will always converge if the data is linearly separable.
	Perceptron algorithm is using the gradient of L2 loss to learn w
	If y(n)y (n) is positive, it means instance n is correctly classified
	Perceptron algorithm is guaranteed to converge when the learning rate is set to 1