Perceptrons

• inputs a vector of real-valued inputs, calculates a linear combination and outputs a 1 if the results is greater than some threshold and -1 otherwise
• 
• each is a real-valued weight that determines the contribution of the input to the perceptron output
• the quantity ( ) is the threshold
• hyperplane decision surface in the n-dimensional space of instances
• not all datasets can be separated by a hyperplane, but if they can they are linearly separableБе datasets