Analysis of learning algorithms for the similarity neural network

thumbnail

Student

Mallela, Pravallika

Document type

Bachelor thesis

Date

2020

rights

Open AccessOpen Access

Publisher

Universitat Politècnica de Catalunya



Abstract

A similarity neural network is a two-layer neural network based on similarity measures: the first layer computes the similarity between the input data and a set of prototypes, and the second layer gathers these results and predicts the output. The goal of the project is to analyze fast training algorithms and compare them with more traditional learning algorithms, such as a standard two-layer MLP network. Along with the usage of similarity measures, this research also aims to develop a model that can handle complex, heterogeneous, imprecise, missing, and complex data without changing their original form.

Entitat col·laboradora

SASTRA Deemed University
user

Participating teacher

Files