Statistical learning Methods Reading notes: Perception machines

What is a perception machine?

The linear classification model of the two classifications, the input is the characteristic vector of the instance, the output is the class of the instance, 1 and 12 values are taken.

The perceptual machine corresponds to the separation of the instances into positive and negative two classes in the input space (feature space), which belongs to the discriminant model.

Perceptron is the basis of neural networks and support vector machines.

Conditions of use of the model

The data set needs to be linearly divided, and if the data set is linear, a perceptron model can be obtained after a certain number of iterations, leaving the positive and negative examples separated.

Definition of loss function

A natural choice is the total number of false classification points, but such loss function is not a continuous derivative of the parameters W and B, and is not easy to optimize.

The loss function used by the Perceptron is the total distance from the error classification point to the super plane S.

Learning Algorithms

The Perceptual machine learning problem is transformed into the optimization problem of solving the loss function.

The optimal method is a random gradient descent.

(1) Arbitrary selection of a super-planar w0 and B0;

(2) The gradient descent continuously minimizes the objective function, and at each time the minimization process is randomly selected by a random classification point to reduce its gradient.

Statistical learning Methods Reading notes: Perception machines