We introduce a variant of the perceptron algorithm called second-order perceptron algorithm, which is able to exploit certain spectral properties of the data. We analyze the second-order perceptron algorithm in the mistake bound model of on-line learning and prove bounds in terms of the eigenvalues of the Gram matrix created from the data. The performance of the second-order perceptron algorithm is affected by the setting of a parameter controlling the sensitivity to the distribution of the eigenvalues of the Gram matrix. Since this information is not preliminarly available to on-line algorithms, we also design a refined version of the second-order perceptron algorithm which adaptively sets the value of this parameter. For this second algorithm we are able to prove mistake bounds corresponding to a nearly optimal constant setting of the parameter.
A second-order Perceptron algorithm
GENTILE, CLAUDIO
2002-01-01
Abstract
We introduce a variant of the perceptron algorithm called second-order perceptron algorithm, which is able to exploit certain spectral properties of the data. We analyze the second-order perceptron algorithm in the mistake bound model of on-line learning and prove bounds in terms of the eigenvalues of the Gram matrix created from the data. The performance of the second-order perceptron algorithm is affected by the setting of a parameter controlling the sensitivity to the distribution of the eigenvalues of the Gram matrix. Since this information is not preliminarly available to on-line algorithms, we also design a refined version of the second-order perceptron algorithm which adaptively sets the value of this parameter. For this second algorithm we are able to prove mistake bounds corresponding to a nearly optimal constant setting of the parameter.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
