Tipo di tesi
Tesi di laurea magistrale
Titolo
Development of incremental learning techniques with SVM and MLP neural networks
Dipartimento
INGEGNERIA DELL'INFORMAZIONE
Corso di studi
COMPUTER ENGINEERING
Riassunto (Italiano)
Development of incremental learning solutions for predictive algorithms support using vector machine and Multy Layer Perceptrons neural networks. The goal is to find a way to update predictive models using new data without reusing previously used data. It also explores the possibility of teaching models in use classes of previously unrecognized objects. In the part related to neural networks we try to stem the problem of catastrophic forgetting. The results obtained are encouraging as far as the Support Vector Machine is concerned, while the approach used for neural networks has not brought hoped-for results.
