Today's supply chains face increasing volatility on many fronts. From the shop-floor where machines break and suppliers fail to the boardrooms where unanticipated price inflation erodes profi tability. Turbulence is the new normal. To remain competitive and weather these (daily) storms, supply chains need to move away from an effi ciency mindset towards a resilience mindset. For over a little more than a decade industry and academia have awakened to this reality. Academic literature and case studies show that there is no longer a shortage of resilience strategies and designs. Unfortunately, industry still lacks the tools with which to assess and evaluate the effectiveness of such strategies and designs. Without the ability to quantify the benefi t it is impossible to motivate the cost. This thesis adds one piece to the puzzle of quantifying supply chain vulnerability. Speci fically, it focussed on supply chains within urban areas. It addresses the question: "How does a supply chain's network design (internal con figuration) and its dependence on the underlying road network (external circumstances) make it more or less vulnerable to disruptions of the road network?" Multilayered Complex Network Theory (CNT) held promise as a modelling approach that could capture the complexity of the dependence between a logical supply chain network and the physical road network that underpins it. This approach addressed two research gaps in complex network theory applications. In the supply chain arena CNT applications have reaped many benefi ts but the majority of studies regarded single-layer networks that model only supply chain relations. There were no studies found where the dependence of supply chain layers on underlying physical infrastructure was modelled in a multilayered manner. Road network applications offered many more multilayered applications but these primarily focussed on passenger transport, not freight transport. The first artefact developed in the thesis was a multilayered complex network formulation representing a logical (supply chain) layer placed on a physical (road infrastructure) layer. The individual layers had predefi ned network characteristics and on their own could not hint at the inherent vulnerability that the system as a whole might have. From the multilayered formulation, the collection of shortest paths emerged. This is the collection of all shortest path alternatives within a network. The collection of shortest paths is the unique fingerprint of each multilayered network instance. The key to understanding vulnerability lies within the characteristics of the collection of shortest paths. Three standard supply chain network archetypes were de fined namely the Fully Connected (FC), Single Hub (SH) and Double Hub (DH) archetypes. A sample of 500 theoretical multilayered network instances was generated for each archetype. These theoretical instances were subjected to three link-based progressive targeted disruption simulations to study the vulnerability characteristics of the collection of shortest paths. Two of the simulations used relative link betweenness to prioritise the disruptions while the third used the concept of network skeletons as captured by link salience. The results from these simulations showed that the link betweenness strategies were far more effective than the link salience strategy. From these results three aspects of vulnerability were identifi ed. Redundancy quantifi es the number of alternative shortest paths available to an instance. Overlap measures to what degree the shortest path sets of an instance overlap and have road segments in common. Effi ciency step-change is a measure of the magnitude of the "shock" absorbed by the shortest paths of an instance during a disruption. For each of these aspects one or more metrics were defi ned. This suite of vulnerability metrics is the second artefact produced by the thesis. The design of the artefacts itself, although novel, was not considered research. It is the insights derived during analysis of the artefacts' performance that contributes to the body of knowledge. Link-based progressive random disturbance simulations were used to assess the ability of the vulnerability metrics to quantify supply chain vulnerability. It was found that none of the de fined vulnerability aspects are good stand-alone predictors of vulnerability. The multilayered nature and random disturbance protocol result in vulnerability being more multi-faceted than initially imagined. Nonetheless, the formulation of the multilayered network proved useful and intuitive and even though the vulnerability metrics fail as predictors they still succeed in capturing shortest path phenomena that would lead to vulnerability under non-random protocols. To validate the fi ndings from the theoretical instances, link-based random disturbance simulations were executed on 191 case study instances. These instances were extracted from real-life data in three urban areas in South Africa, namely Gauteng Province (GT), City of Cape Town (CoCT) and eThekwini Metropolitan Municipality (ET). The case study instances showed marked deviations from the assumptions underlying the theoretical instances. Despite these differences, the multilayered formulation still enables the quanti fication of the relationship between supply chain structure and road infrastructure. The performance of the vulnerability metrics in the case study corroborates the findings from the theoretical instances. Although the suite of vulnerability metrics was unsuccessful in quantifying or predicting vulnerability in both the theoretical and case study instances, the rationale behind their development is sound. Future work that will result in more effective metrics is outlined in this thesis. On the one hand the development of a more realistic disruption strategy is suggested. Road network disruptions are neither completely random nor specifi cally targeted. Important segments with greater tra ffic loads are more likely to be disrupted, but the reality is that disruptions such as accidents, equipment failure or road maintenance could really occur anywhere on the network. A more realistic disruption strategy would lie somewhere on the continuum between targeted and random disruptions. Other future work suggests the refi nement of both artefacts by incorporating link weights in both the logical and physical layers. An unanticipated fi nding from this thesis is that future research in the fi eld may be expedited if theory-building emanates from real-life empirical networks as opposed to theoretically generated networks. Expanding the scope of the case study, characterising the true network archetypes found in practice and increasing the number of case study samples is a high priority for future work. / Thesis (PhD)--University of Pretoria, 2018. / National Research Foundation of South Africa (Grant UID: 105519). Partial funding of doctoral research. / Industrial and Systems Engineering / PhD / Unrestricted

Master of Science / Department of Statistics / Perla Reyes / Networks are defined as sets of items and their connections. Interconnected items are represented by mathematical abstractions called vertices (or nodes), and the links connecting pairs of vertices are known as edges. Networks are easily seen in everyday life: a network of friends, the Internet, metabolic or citation networks. The increase of available data and the need to analyze network have resulted in the proliferation of models for networks. However, for networks with billions of nodes and edges, computation and inference might not be achieved within a reasonable amount of time or budget. A sampling approach seems a natural choice, but traditional models assume that we can have access to the entire network. Moreover, when data is only available for a sampled sub-network conclusions tend to be extrapolated to the whole network/population without regard to sampling error. The statistical problem this report addresses is the issue of how to sample a sub-network and then draw conclusions about the whole network. Are some sampling techniques better than others? Are there more efficient ways to estimate parameters of interest? In which way can we measure how effectively my method is reproducing the original network? We explore these questions with a simulation study on Mesa High School students' friendship network. First, to assess the characteristics of the whole network, we applied the traditional exponential random graph model (ERGM) and a stochastic blockmodel to the complete population of 205 students. Then, we drew simple random and stratified samples of 41 students, applied the traditional ERGM and the stochastic blockmodel again, and defined a way to generalized the sample findings to the population friendship network of 205 students. Finally, we used the degree distribution and other network statistics to compare the true friendship network with the projected one. We achieved the following important results: 1) as expected stratified sampling outperforms simple random sampling when selecting nodes; 2) ERGM without restrictions offers a poor estimate for most of the tested parameters; and 3) the Bayesian stochastic blockmodel estimation using a strati ed sample of nodes achieves the best results.

The explosion of personal positioning devices like GPS-enabled smartphones has enabled the collection and storage of a huge amount of positioning data in the form of trajectories. Thereby, trajectory data have brought many research challenges in the process of recovery, storage and knowledge discovery in mobility as well as new applications to support our society in mobility terms. Other research area that has been receiving great attention nowadays is the area of complex network or science of networks. Complex network is the first approach to model complex system that are present in the real world, such as economic markets, the Internet, World Wide Web and disease spreading to name a few. It has been applied in different field, like Computer Science, Biology and Physics. Therefore, complex networks have demonstrated a great potential to investigate the behavior of complex systems through their entities and the relationships that exist among them. The present dissertation, therefore, aims at exploiting approaches to analyze mobility data using a perspective of complex networks. The first exploited approach stands for the trajectories as the main entities of the networks connecting each other through a similarity function. The second, in turn, focuses on points of interest that are visited by people, which perform some activities in these points. In addition, this dissertation also exploits the proposed methodologies in order to develop a software tool to support users in mobility analysis using complex network techniques.

A vasta quantidade de informações disponível sobre qualquer área de conhecimento torna cada vez mais difícil selecionar e analisar informações específicas e relevantes sobre determinado assunto. Com isso, faz-se necessário o aprimoramento de técnicas automáticas para recuperação, análise e extração de conhecimento em conjuntos de dados, destacando-se dessa forma as pesquisas em Aprendizado de Máquina e em Mineração de Dados. Em aprendizado de máquina e em mineração, a grande maioria das técnicas utiliza-se de uma representação proposicional dos dados, que considera apenas caracter características individuais dos objetos descritos em uma tabela atributo-valor. Porém, existem aplicações nas quais além da descrição dos objetos também estão disponíveis informações sobre relações existentes entre eles. Esses domínios podem ser representados via grafos, nos quais vértices representam objetos e arestas relações entre objetos, possibilitando a aplicação de técnicas relacionais aos dados. Conceitos de Redes Complexas (RC) podem ser utilizados neste contexto. RC é um campo de pesquisa recente e ativo, que estuda o comportamento de diversos sistemas reais, modelados via grafos. Entretanto, ainda há poucos trabalhos que utilizam Redes Complexas em aprendizado de máquina ou mineração de dados. Este projeto apresenta uma proposta de utilização do formalismo de redes complexas e grafos para descoberta de padrões no contexto de aprendizado supervisionado. O formalismo de grafos permite representar as relações entre objetos e características particulares do domínio, permitindo agregar informações estruturais das relações à descoberta de conhecimento. Especificamente, neste trabalho desenvolve-se uma representação relacional baseada em grafos construídos a partir de relações de similaridade entre objetos. Baseado nesta representação são propostas abordagens de classificação relacional. Também é proposto um modelo de rede denominado K-Associados. Propriedades da rede K-Associados foram investigadas. Os resultados experimentais demonstram um grande potencial para classificação utilizando os algoritmos de classificação e de formação de redes propostos / The vast amount of information available on any area of knowledge makes selecting and analyzing information on a specific topic increasingly dificult. Therefore, it is necessary the improvement of techniques for automatic information retrieval, analysis, and knowledge extraction from data sets. In this scenario, especial attention must be addressed for Machine Learning and Data Mining researches. In machine learning and data mining, most of the techniques uses a propositional representation, which considers only the characteristics of the objects described into an attribute-value table. However, there are domains where, in addition to the description of the objects, it is also available information about relationship between them. Such domains can be represented by graphs where vertices represent objects and edges relationship between objects, enabling the application of techniques for relational data. Concepts of complex networks (CN) can be useful in this context. CN is a recent and active research field, which studies the behavior of many real systems modeled by graphs. However, there is little work in machine learning or data mining applying CN concepts. This project presents a proposal to use the formalism of complex networks and graphs to discover patterns in the context of supervised learning. The formalism of graphs can represent relationships between objects and characteristics of the domain, allowing adding structural knowledge embedded in a graph into the data mining process. Specifically, this work develops a relational representation based on graphs constructed taking into consideration the similarity between objects. Based on this representation, relational classification approaches are proposed. It is also proposed a network referred to K-Associate Network. Properties of the K-Associate Network were investigated. The experimental results show great potential for the proposed classification and network construction algorithms

It is often a challenge to atomically determine the structure of large macromolecular assemblies, even if successfully crystallized, due to their weak diffraction of X-rays. Refinement algorithms that work with low-resolution diffraction data are necessary for researchers to obtain a picture of the structure from limited experimental information. Relationship between the structure and function of proteins implies that a refinement approach delivering accurate structures could considerably facilitate further research on their function and other related applications such as drug design. Here a refinement algorithm called the Deformable Complex Network is presented. Computation results revealed that, significant improvement was observed over the conventional refinement and DEN refinement, across a wide range of test systems from the Protein Data Bank, indicated by multiple criteria, including the free R value, the Ramachandran Statistics, the GDT (<1Å) score, TM-score as well as associated electron density map.

A vasta quantidade de informações disponível sobre qualquer área de conhecimento torna cada vez mais difícil selecionar e analisar informações específicas e relevantes sobre determinado assunto. Com isso, faz-se necessário o aprimoramento de técnicas automáticas para recuperação, análise e extração de conhecimento em conjuntos de dados, destacando-se dessa forma as pesquisas em Aprendizado de Máquina e em Mineração de Dados. Em aprendizado de máquina e em mineração, a grande maioria das técnicas utiliza-se de uma representação proposicional dos dados, que considera apenas caracter características individuais dos objetos descritos em uma tabela atributo-valor. Porém, existem aplicações nas quais além da descrição dos objetos também estão disponíveis informações sobre relações existentes entre eles. Esses domínios podem ser representados via grafos, nos quais vértices representam objetos e arestas relações entre objetos, possibilitando a aplicação de técnicas relacionais aos dados. Conceitos de Redes Complexas (RC) podem ser utilizados neste contexto. RC é um campo de pesquisa recente e ativo, que estuda o comportamento de diversos sistemas reais, modelados via grafos. Entretanto, ainda há poucos trabalhos que utilizam Redes Complexas em aprendizado de máquina ou mineração de dados. Este projeto apresenta uma proposta de utilização do formalismo de redes complexas e grafos para descoberta de padrões no contexto de aprendizado supervisionado. O formalismo de grafos permite representar as relações entre objetos e características particulares do domínio, permitindo agregar informações estruturais das relações à descoberta de conhecimento. Especificamente, neste trabalho desenvolve-se uma representação relacional baseada em grafos construídos a partir de relações de similaridade entre objetos. Baseado nesta representação são propostas abordagens de classificação relacional. Também é proposto um modelo de rede denominado K-Associados. Propriedades da rede K-Associados foram investigadas. Os resultados experimentais demonstram um grande potencial para classificação utilizando os algoritmos de classificação e de formação de redes propostos / The vast amount of information available on any area of knowledge makes selecting and analyzing information on a specific topic increasingly dificult. Therefore, it is necessary the improvement of techniques for automatic information retrieval, analysis, and knowledge extraction from data sets. In this scenario, especial attention must be addressed for Machine Learning and Data Mining researches. In machine learning and data mining, most of the techniques uses a propositional representation, which considers only the characteristics of the objects described into an attribute-value table. However, there are domains where, in addition to the description of the objects, it is also available information about relationship between them. Such domains can be represented by graphs where vertices represent objects and edges relationship between objects, enabling the application of techniques for relational data. Concepts of complex networks (CN) can be useful in this context. CN is a recent and active research field, which studies the behavior of many real systems modeled by graphs. However, there is little work in machine learning or data mining applying CN concepts. This project presents a proposal to use the formalism of complex networks and graphs to discover patterns in the context of supervised learning. The formalism of graphs can represent relationships between objects and characteristics of the domain, allowing adding structural knowledge embedded in a graph into the data mining process. Specifically, this work develops a relational representation based on graphs constructed taking into consideration the similarity between objects. Based on this representation, relational classification approaches are proposed. It is also proposed a network referred to K-Associate Network. Properties of the K-Associate Network were investigated. The experimental results show great potential for the proposed classification and network construction algorithms

Highway bridges are among the most vulnerable and expensive components in transportation networks. In response, the Government of Ontario has allocated $26 billion in the next 10 years to address issues pertaining to aging bridge and deteriorating highway infrastructure in the province. Although several approaches have been developed to guide their rehabilitation, most bridge rehabilitation approaches are focused on the component level (individual bridge) in a relative isolation of other bridges in the network. The current study utilizes a complex network theoretic approach to quantify the topological characteristics of the Ontario Bridge Network (OBN) and subsequently evaluate the OBN robustness and vulnerability characteristics. These measures are then integrated in the development of a Multi Scale Bridge Classification (MSBC) approach—an innovative classification approach that links the OBN component level data (i.e., Bridge Condition Index and year of construction, etc.) to the corresponding dynamic network-level measures. The novel approach calls for a paradigm shift in the strategy governing classifying and prioritizing bridge rehabilitation projects based on bridge criticality within the entire network, rather than only the individual bridge’s structural conditions. The model was also used to identify the most critical bridges in the OBN under different disruptions to facilitate rapid implementation of the study results. / Thesis / Master of Applied Science (MASc)

No / Symptoms and signs (symptoms in brief) are the essential clinical manifestations for traditional Chinese medicine (TCM) diagnosis and treatments. To gain insights into the molecular mechanism of symptoms, this paper presents a network-based data mining method to integrate multiple phenotype-genotype data sources and predict the prioritizing gene rank list of symptoms. The result of this pilot study suggested some insights on the molecular mechanism of symptoms.

No / Traditional Chinese medicine (TCM) investigates the clinical diagnosis and treatment regularities in a typical schema of personalized medicine, which means that individualized patients with same diseases would obtain distinct diagnosis and optimal treatment from different TCM physicians. This principle has been recognized and adhered by TCM clinical practitioners for thousands of years. However, the underlying mechanisms of TCM personalized medicine are not fully investigated so far and remained unknown. This paper discusses framework of TCM personalized medicine in classic literatures and in real-world clinical settings, and investigates the underlying mechanisms of TCM personalized medicine from the perspectives of network medicine. Based on 246 well-designed outpatient records on insomnia, by evaluating the personal biases of manifestation observation and preferences of herb prescriptions, we noted significant similarities between each herb prescriptions and symptom similarities between each encounters. To investigate the underlying mechanisms of TCM personalized medicine, we constructed a clinical phenotype network (CPN), in which the clinical phenotype entities like symptoms and diagnoses are presented as nodes and the correlation between these entities as links. This CPN is used to investigate the promiscuous boundary of syndromes and the co-occurrence of symptoms. The small-world topological characteristics are noted in the CPN with high clustering structures, which provide insight on the rationality of TCM personalized diagnosis and treatment. The investigation on this network would help us to gain understanding on the underlying mechanism of TCM personalized medicine and would propose a new perspective for the refinement of the TCM individualized clinical skills.

Yes / As a global strategic reserve resource, rare earth has been widely used in important industries, such as military equipment and biomedicine. However, through existing analyses based on the total volume of rare earth trade, the competition and dependency behind the trade cannot be revealed. In this paper, based on the principle of trade preference and import similarity, we construct dependency and competition networks and use complex network analysis to study the evolution of the global rare earth trade network from 2002 to 2018. The main conclusions are as follows: the global rare earth trade follows the Pareto principle, and the trade network shows a scale-free distribution. China has become the largest country in both import and export of rare earth trade in the world since 2017. In the dependency network, China has become the most dependent country since 2006. The result of community division shows that China has separated from the American community and formed new communities with the Association of Southeast Asian Nations (ASEAN) countries. The United States of America has formed a super-strong community with European and Asian countries. In the competition network, the distribution of competition intensity follows a scale-free distribution. Most countries are faced with low-intensity competition, but competing countries are relatively numerous. The competition related to China has increased significantly. The competition source of the United States of America has shifted from Mexico to China. China, the USA, and Japan have been the cores of the competition network. / This work was supported by the Ministry of Education of the People’s Republic of China Humanities and Social Sciences Youth Foundation (Grant No. 22YJC910014), the Social Sciences Planning Youth Project of Anhui Province (Grant No. AHSKQ2022D138), and the Innovation Development Research Project of Anhui Province (Grant No. 2021CX053).