Recently, several works have approached the HIV-1 protease specificity problem by applying a number of classifier creation and combination methods, from the field of machine learning. In this work we propose a hierarchical classifier (HC) architecture. We confirm previous results stating that linear classifiers obtain higher performance than non-linear classifier using orthonormal encoding. Moreover, we prove that our new hierarchical approach is a successful attempt to obtain a drastically error reduction with respect to the performance of linear classifiers. The error rate decreases from 9.1% using linear support vector machines to 7% using the new hierarchical classifier
Titolo: | Machine Learning for HIV-1 Protease Cleavage Site Prediction |
Autore/i: | LUMINI, ALESSANDRA; NANNI, LORIS |
Autore/i Unibo: | |
Anno: | 2005 |
Titolo del libro: | Artificial Intelligence and Applications (AIA 2005) |
Pagina iniziale: | 406 |
Pagina finale: | 410 |
Abstract: | Recently, several works have approached the HIV-1 protease specificity problem by applying a number of classifier creation and combination methods, from the field of machine learning. In this work we propose a hierarchical classifier (HC) architecture. We confirm previous results stating that linear classifiers obtain higher performance than non-linear classifier using orthonormal encoding. Moreover, we prove that our new hierarchical approach is a successful attempt to obtain a drastically error reduction with respect to the performance of linear classifiers. The error rate decreases from 9.1% using linear support vector machines to 7% using the new hierarchical classifier |
Data prodotto definitivo in UGOV: | 28-set-2005 |
Appare nelle tipologie: | 4.01 Contributo in Atti di convegno |