Tesi di Dottorato

Permanent URI for this communityTesi di Dottorato

Browse

Author: search.filters.author.Grandinetti, LucioSubject: search.filters.subject.Algoritmi

Search Results

Now showing 1 - 6 of 6
  • No Thumbnail Available
    Item
    Simulation-based optimization in port logistics
    (2017) Mazza, Rina Mary; Grandinetti, Lucio; Legato, Pasquale
  • No Thumbnail Available
    Item
  • No Thumbnail Available
    Item
    A branch and cut approach for the mixed capacitated general routing problem
    (2014-06-06) Bosco, Adamo; Laganà, Demetrio; Grandinetti, Lucio
    The issue addressed in this thesis consists in modeling and solving the Mixed Capacitated General Routing Problem (MCGRP). This problem generalizes many routing problems, either in the Node or in the Arc routing family. This makes the problem a very general one and gives it a big interest in realworld applications. Despite this, and because of the native di culty of the problem, very few papers have been devoted to this argument. In the thesis an integer programming model for the MCGRP is proposed and several valid inequalities for the undirected Capacitated Arc Routing polyhedron are extended and generalized to the MCGR polyhedron. A branch and cut algorithm for the MCGRP is developed and tested on a dataset of new instances derived from mixed CARP benchmark instances. Moreover an heuristic procedure is de ned in order to nd a good upper bound aimed at helping the branch and cut algorithm to cut o unpromising regions of the search tree. This schema will be used and extended in future works to solve bigger real-world instances. Extensive numerical experiments indicate that the algorithm is able to optimally solve many instances. In general, it provides valid lower and upper bounds for the problem in a reasonable amount of time.
  • No Thumbnail Available
    Item
    Topics in real-time fleet management
    (2014-06-06) Manni, Emanuele; Grandinetti, Lucio; Ghiani, Gianpaolo; Barrett, W. Thomas
  • No Thumbnail Available
    Item
    Un framework di soluzione ad alto livello per problemi di classificazione basato su approcci metaeuristici
    (2014-05-27) Candelieri, Antonio; Grandinetti, Lucio; Conforti, Domenico
    This work deals with the development and implementation of a high-level classification framework which combines parameters optimization of a single classifier with classifiers ensemble optimization, through meta-heuristics. Support Vector Machines (SVM) is used for learning while the meta-heuristics adopted and compared are Genetic-Algorithms (GA), Tabu-Search (TS) and Ant Colony Optimization (ACO). Single SVM optimization usually concerns two approaches: searching for optimal parameter values of a SVM with a fixed kernel (Model Selection) or with a linear combination of basic kernels (Multiple Kernel Learning), both approaches have been taken into account. Adopting meta-heuristics avoids to perform time consuming grid-approach for testing several classifier configurations. In particular, starting from canonical formulation of GA, this study proposes some changes in order to take into account specificities of classification learning. Proposed solution has been extensively tested on 8 classification datasets (5 of them are of public domain) providing reliable solutions and showing to be effective. In details, unifying Model Selection, Multiple Kernel Learning and Ensemble Learning on a single framework proved to be a comprehensive and reliable approach, and showing that best solutions have been identified by one of the strategies depending on decision problem and/or available data. Under this respect, the proposed framework may represent a new effective and efficient high-level SVM classification learning strategy.
  • No Thumbnail Available
    Item
    Omega our multi ethnic genetic algorithm
    (2014-03-13) Cerrone, Carmine; Grandinetti, Lucio; Gaudioso, Manlio
    Combinatorial optimization is a branch of optimization. Its domain is optimization problems where the set of feasible solutions is discrete or can be reduced to a discrete one, the goal being that of nding the best possible solution. Two fundamental aims in optimization are nding algorithms characterized by both provably good run times and provably good or even optimal solution quality. When no method to nd an optimal solution, under the given constraints (of time, space etc.) is available, heuristic approaches are typically used. A metaheuristic is a heuristic method for solving a very general class of computational problems by combining user- given black-box procedures, usually heuristics themselves, in the hope of obtaining a more e cient or more robust procedure. The genetic algorithms are one of the best metaheuristic approaches to deal with optimization problems. They are a population- based search technique that uses an ever changing neighborhood structure, based on population evolution and genetic operators, to take into account di erent points in the search space. The core of the thesis is to introduce a variant of the classic GA approach, which is referred to as OMEGA (Multi Ethnic Genetic Algorithm). The main feature of this new metaheuristic is the presence of di erent populations that evolve simultaneously, and exchange genetic material with each other. We focus our attention on four di erent optimization problems de ned on graphs. Each one is iii iv proved to be NP-HARD. We analyze each problem from di erent points of view, and for each one we de ne and implement both a genetic algorithm and our OMEGA.