Suppose you are given a dataset with four points and two classes. First plot these points for yourself, and convince yourself that any linear classifier will be unable to classify all 4 points correctly. We have drawn you a simple 1 hidden layer MLP, with the 2 inputs (X1, X2), 2 hidden nodes (H1, H2), 2 biases (B1, B2) and one output. All of the parameters are written in the table as wel as illustrated on the figure where Wi is there edge weight that connects node i in layer L to node j in layer L+1. Assume that the sign activation function (i.e. returning +1 if input is positive, and -1 otherwise) is used at each node, except for the output node. For the output node, assume the logistic function is used to produce P(y - 1lx). Give weights for this MLP that will perfect classify the training set X1 X2 Y WeightValue 0 WI 12 21 Out lyl 12 WI 12 VI 2 /2 12 Suppose you are given a dataset with four points and two classes. First plot these points for yourself, and convince yourself that any linear classifier will be unable to classify all 4 points correctly. We have drawn you a simple 1 hidden layer MLP, with the 2 inputs (X1, X2), 2 hidden nodes (H1, H2), 2 biases (B1, B2) and one output. All of the parameters are written in the table as wel as illustrated on the figure where Wi is there edge weight that connects node i in layer L to node j in layer L+1. Assume that the sign activation function (i.e. returning +1 if input is positive, and -1 otherwise) is used at each node, except for the output node. For the output node, assume the logistic function is used to produce P(y - 1lx). Give weights for this MLP that will perfect classify the training set X1 X2 Y WeightValue 0 WI 12 21 Out lyl 12 WI 12 VI 2 /2 12