Submitted by Marlene Santos (marlene.bc.ufg@gmail.com) on 2018-11-05T18:20:10Z No. of bitstreams: 2 Dissertação - Welton Cardoso do Carmo - 2018.pdf: 38319203 bytes, checksum: 38a0892d45bc6618a1051e2785d397ca (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2018-11-06T10:01:23Z (GMT) No. of bitstreams: 2 Dissertação - Welton Cardoso do Carmo - 2018.pdf: 38319203 bytes, checksum: 38a0892d45bc6618a1051e2785d397ca (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2018-11-06T10:01:23Z (GMT). No. of bitstreams: 2 Dissertação - Welton Cardoso do Carmo - 2018.pdf: 38319203 bytes, checksum: 38a0892d45bc6618a1051e2785d397ca (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2018-10-01 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / he metropolises face problems in the collective transport system with buses, whether in the stocking of vehicles, cost of tickets or number of vehicles in circulation, which influence the quality of mobility services provided. Before proposing changes in the system in order to solve some problems, it is necessary to find a way to measure such quality, in order to have a basis for the study / comparison between the current system and a modified one. Thus, the present work describes an adaptation of a traditional model of the collective transportation system, with the bus as the main means of transportation. The adapted model has the objective of maximizing the quality of bus transportation and, at the same time, allowing simulations of some system scenarios to be performed. A simulator with the intention of testing such modeling is presented. Finally, to verify the effectiveness of the model and the simulator, the data of the collective transportation system of the city of Goiânia, Goiás, are used. / As metrópoles enfrentam problemas no sistema de transporte coletivo com ônibus, seja na lotação dos veículos, custo das passagens ou número de veículos em circulação, que influenciam na qualidade dos serviços de mobilidade prestados. Antes de propor alterações no sistema com o intuito de resolver alguns problemas, é necessário encontrar uma forma de medir tal qualidade, a fim de se ter uma base para o estudo/comparação entre o sistema atual e um modificado. Assim, o presente trabalho descreve uma adaptação de um modelo tradicional do sistema de transporte coletivo, tendo o ônibus como principal meio de transporte. A modelagem adaptada tem como objetivo maximizar a qualidade do transporte com ônibus e, ao mesmo tempo, permitir que simulações de alguns cenários do sistema possam ser realizadas. Um simulador com o intuito de testar tal modelagem é apresentado. Por fim, para verificar a eficácia do modelo e do simulador, são utilizados os dados do sistema de transporte coletivo da cidade de Goiânia, Goiás.

This master's thesis deals with optimization of traffic networks. There are discussed modeling methods for traffic and simplifications used in these models. Introductory part is focused on mathematical theory used to buid presented model. Main focus of this thesis was creation and optimization of the model which describes real world traffic network with traffic lights. Model was tested on both artificial and real data gathered in city Zlín. It was necessary to create generator which can produce suitable input data for model.

Online marketing has resulted in a paradigm shift in the advertisement industry. Programmatic buying is an emerging business model that is very promising for online advertising. In online advertising, revenue maximization is always a key matter for publishers. The purpose of this report is to examine whether programmatic buying can be used in conjunction with other parameters such as Google Adwords or Google Trends to increase profit. This research will provide valuable information regarding how to obtain site visitors at a cheap price while maximizing profit on advertisement shown to those users. We investigate a revenue maximization model that calculates the popularity of a set of news in different countries and compares it to the CPM, Cost-Per-Mille, of the corresponding country. To calculate the popularity, the program uses an API from Google Trends and the CPM data is obtained from the company Adform. Furthermore, we originally planned to also include Google Adwords to estimate the price of traffic acquisition. However, since we found several problems with achieving reasonable estimates for our purpose this parameter has therefore been excluded from the final product. The final product can therefore be seen as a soft indicator of how popular different news are in different countries and what revenue can be expected from corresponding countries. / Digital marknadsföring har resulterat i ett paradigmskifte inom reklambranschen. Programmatisk handel är en mycket lovande affärsmodell för automatisk annonsering online och är under stark tillväxt. Inom digital marknadsföring är vinstmaximering alltid en nyckelfråga för utgivare av annonsplatser. Denna rapport ämnar undersöka huruvida programmatisk handel kan användas tillsammans med andra parametrar som Google Adwords eller Google Trends för att öka vinsten från video-reklamannonser. Den grundläggande idén är att skapa trafik till en specifik hemsida för ett så lågt pris som möjligt samtidigt som vi vill att reklamvisningarna ska ge så höga intäkter som möjligt. Rapporten utreder parametrar som videopris för programmatisk handel i olika länder, Bounce Rate, Cost Per Click och Google Trends Score. Dessa parametrar används för att skapa en sammanvägning för att indikera i vilka länder och för vilka sökord den ekonomiska vinsten potentiellt är störst.        Arbetet har resulterat i ett program som beräknar populariteten för ett antal nyheter i olika länder och jämför med CPM, Cost-Per-Mille, priset för motsvarande land. För att beräkna populariteten används ett API från Google Trends och CPM datan kommer från företaget Adform. Från början var tanken att även väga in Google Adwords för att skapa en prisbild över kostnaden att inbringa trafik. Begränsningar som behövt genomföras under arbetets gång är att exkludera Google Adwords prissättning i det färdiga programmet, då det finns svårigheter i att utröna exakta prisuppgifter från Google Adwords. Slutprodukten är därmed en indikator på vilka nyheter som är populära i olika länder och intäkterna som kan förväntas därifrån.

Sub-optimal automated transportation control systems incur high mobility, human health and environmental costs. With society reliant on its transportation systems for the movement of individuals, goods and services, minimizing these costs benefits many. Intersection traffic signal controllers are an important element of modern transportation systems that govern how vehicles traverse road infrastructure. Many types of traffic signal controllers exist; fixed time, actuated and adaptive. Adaptive traffic signal controllers seek to minimize transportation costs through dynamic control of the intersection. However, many existing adaptive traffic signal controllers rely on heuristic or expert knowledge and were not originally designed for scalability or for transportation’s big data future. This research addresses the aforementioned challenges by developing a scalable system for adaptive traffic signal control model development using deep reinforcement learning in traffic simulation. Traffic signal control can be modelled as a sequential decision-making problem; reinforcement learning can solve sequential decision-making problems by learning an optimal policy. Deep reinforcement learning makes use of deep neural networks, powerful function approximators which benefit from large amounts of data. Distributed, parallel computing techniques are used to provide scalability, with the proposed methods validated on a simulation of the City of Luxembourg, Luxembourg, consisting of 196 intersections. This research contributes to the body of knowledge by successfully developing a scalable system for adaptive traffic signal control model development and validating it on the largest traffic microsimulator in the literature. The proposed system reduces delay, queues, vehicle stopped time and travel time compared to conventional traffic signal controllers. Findings from this research include that using reinforcement learning methods which explicitly develop the policy offers improved performance over purely value-based methods. The developed methods are expected to mitigate the problems caused by sub-optimal automated transportation signal controls systems, improving mobility and human health and reducing environmental costs. / Thesis / Doctor of Philosophy (PhD) / Inefficient transportation systems negatively impact mobility, human health and the environment. The goal of this research is to mitigate these negative impacts by improving automated transportation control systems, specifically intersection traffic signal controllers. This research presents a system for developing adaptive traffic signal controllers that can efficiently scale to the size of cities by using machine learning and parallel computation techniques. The proposed system is validated by developing adaptive traffic signal controllers for 196 intersections in a simulation of the City of Luxembourg, Luxembourg, successfully reducing delay, queues, vehicle stopped time and travel time.