The issues of performance measurement and efficiency analyses for transit industries have been gaining significance due to severe operating conditions and financial constraints in which these transit agencies provide service. In this research, we present an approach to measure the performance of Urban Transit Networks, specifically, bus routes that comprise the network. We propose a math programming model that evaluates the efficiencies of bus routes taking into consideration, the service providers, the users and the societal perspectives. This model is based on Data Envelopment Analysis (DEA) methodology and derives from Network Theory, Network Modeling in DEA, Goal Programming & Goal-DEA and 'Environmental' Variables. This approach enables the decision maker to determine the performance of its units of operations ('bus routes' in our case), optimally allocate scarce resources and achieve target levels for 'externality' variables for these bus routes and for the whole network. We further recommend modifications to the model, for adaptation to other modes of transportation as well as extend its applicability to other applications/scenarios. / Master of Science

Evaluation of critical transportation infrastructure and their operation is vital for continuous evolution to meet the growing needs of the society with time. The current practice of evaluating signalized intersections has two steps. The first is to determine the level of service at which the intersection is performing. Level of Service (LOS) is based on the average delay per vehicle that gets past the particular intersection under consideration. The second step is to do a capacity analysis. This considers the number of lanes and other infrastructure related factors and also includes the influence of the control strategies. The above-described procedure evaluates any one intersection at a time. It is necessary to compare and rank a given set of intersections for planning purposes such as choosing the sites for improvements. The research work presented in this thesis demonstrates how Data Envelopment Analysis (DEA) can be used as a tool to achieve the purpose of comparing and ranking a given set of comparable intersections. This study elaborates on various ways of representing different characteristics of an intersection. The demonstration has been restricted to four way signalized intersections. The intersections that were used for demonstration as part of this research were created in a controlled random fashion by simulation. / Master of Science

Yes / Electronic government services (e-services) involve the delivery of information and services to stakeholders via the Internet, Internet of Things and other traditional modes. Despite their beneficial values, the overall level of usage (take-up) remains relatively low compared to traditional modes. They are also challenging to evaluate due to behavioral, economical, political, and technical aspects. The literature lacks a methodology framework to guide the government transformation application to improve both internal processes of e-services and institutional transformation to advance relationships with stakeholders. This paper proposes a cognitive analytics management (CAM) framework to implement such transformations. The ambition is to increase users’ take-up rate and satisfaction, and create sustainable shared values through provision of improved e-services. The CAM framework uses cognition to understand and frame the transformation challenge into analytics terms. Analytics insights for improvements are generated using Data Envelopment Analysis (DEA). A classification and regression tree is then applied to DEA results to identify characteristics of satisfaction to advance relationships. The importance of senior management is highlighted for setting strategic goals and providing various executive supports. The CAM application for the transforming Turkish e-services is validated on a large sample data using online survey. The results are discussed; the outcomes and impacts are reported in terms of estimated savings of more than fifteen billion dollars over a ten-year period and increased usage of improved new e-services. We conclude with future research.

Yes / Over the past few decades, the banking sectors in Latin America have undergone rapid structural changes to improve the efficiency and resilience of their financial systems. The up-to-date literature shows that all the research studies conducted to analyze the above-mentioned efficiency are based on a deterministic data envelopment analysis (DEA) model or econometric frontier approach. Nevertheless, the deterministic DEA model suffers from a possible lack of statistical power, especially in a small sample. As such, the current research paper develops the technique of satisficing DEA to examine the still less explored case of Peru. We propose a Satisficing DEA model applied to 14 banks operating in Peru to evaluate the bank-level efficiency under a stochastic environment, which is free from any theoretical distributional assumption. The proposed model does not only report the bank efficiency, but also proposes a new framework for peer mining based on the Bayesian analysis and potential improvements with the bias-corrected and accelerated confidence interval. Our study is the first of its kind in the literature to perform a peer analysis based on a probabilistic approach.

No / The quest for best practices may lead to an increased risk of poor decision-making, especially when aiming to attain best practice levels reveals that efforts are beyond the organization’s present capabilities. This situation is commonly known as the “best practice trap”. Motivated by such observation, the purpose of the present paper is to develop a practical methodology to support better practice benchmarking, with an application to the banking sector. In this sense, we develop a two-stage hybrid model that employs Artificial Neural Network (ANN) via integration with Data Envelopment Analysis (DEA), which is used as a preprocessor, to investigate the ability of the DEA-ANN approach to classify the sampled branches of a Greek bank into predefined efficiency classes. ANN is integrated with a family of radial and non-radial DEA models. This combined approach effectively captures the information contained in the characteristics of the sampled branches, and subsequently demonstrates a satisfactory classification ability especially for the efficient branches. Our prediction results are presented using four performance measures (hit rates): percent success rate of classifying a bank branch’s performance exactly or within one class of its actual performance, as well as just one class above the actual class and just one class below the actual class. The proposed modeling approach integrates the DEA context with ANN and advances benchmarking practices to enhance the decision-making process for efficiency improvement.

Yes / The current study contributes to the literature in efficiency analysis in two ways: 1) we build on the existing studies in Dynamic Network Data Envelopment Analysis (DNDEA) by proposing a sequential structure incorporating dual-role characteristics of the production factors; 2) we initiate the efforts to complement the proposal of our innovative sequential DNDEA through a behavioural-causal analysis. The proposal of this statistical analysis is very important considering it does not only validate the proposal of the efficiency analysis but also our practice can be generalized to the future studies dealing with designing innovative production process. Finally, we apply these two different analyses to the banking industry. Using a sample of 43 Chinese commercial banks including five different ownership types (state-owned, joint-stock, city, rural, and foreign banks) between 2010 and 2018, we find that the inefficiency level is around 0.14, although slight volatility has been observed. We find that the highest efficiency is dominated by state-owned banks, and although foreign banks are less efficient than joint-stock banks, they are more efficient than city banks. Finally, we find that rural banks have the highest inefficiency.

Yes / For meeting the external requirements of the Paris Agreement and reducing energy consumption per gross domestic product, China needs to improve its energy efficiency. Although the existing studies have attempted to investigate energy efficiency from different perspectives, little effort has yet been made to consider the collaboration among different stages in the production chain to produce energy outputs. In addition, various studies have also examined the determinants of energy efficiency, however, they mainly focused on technology and economic factors, no study has yet proposed and considered the influence of geographical factors on energy efficiency. In this article, we fill in the gap and make theoretical and empirical contributions to the literature. In this study, a two-stage analysis method is used to analyse energy efficiency and the influencing factors in China between 2009 and 2021. More specifically, from the theoretical/methodological perspective, a multi-activity network data envelopment analysis model is used to measure energy efficiency of different processes in the energy production chain. From the empirical perspective, we attempt to investigate the influence of geographical factors on energy efficiency through a neural network analysis. Meanwhile, the comparisons among different provinces are made. The result shows that the overall energy efficiency is low in China, and China relies more on the traditional energy industry than the clean energy industry. The efficiency level experiences a level of volatility over the examined period. Finally, we find that raw fuel pre-process and industry have a significant and positive impact on energy efficiency in China.

Yes / Using a sample of Brazilian listed companies during 2010-2019, the study investigates the endogeneity and the directional cause-effect relationship between firm efficiency, market structure and firms’ ESG performance under a Stochastic Structural Relationship Programming (SSRP) model. Also, comprehensive market structure indicators are used. The efficiency is estimated under a two-stage network Data Envelopment Analysis (NDEA) model. Our empirical evidence is threefold. First, our evidence indicates that firms with better environmental performance are more efficient, whereas lower ESG performance and poorer corporate governance practices are associated with a higher level of efficiency. Second, our findings suggest that market structure measures (i.e., competition and market power) have heterogeneous impacts on various ESG indexes. Specifically, higher market competition is associated with better overall ESG performance and environmental performance but worse corporate governance performance, although market power can only enhance the environmental and governance performance of firms. Third, the two market structure proxies employed in this study are significantly attributed to firm efficiency. Our findings provide practical implications for various stakeholders and suggest avenues for future studies that can build on our evidence.

Yes / This study estimates output market power in the Chinese banking industry using the multi-output Lerner index. We propose a minimum distance cost function approach, which allows us to determine not only the level of market power but also the non-profit maximizers and efficiency level of Chinese banks. Following the first-stage analysis, we employ the generalized method of moment system estimator to evaluate the impacts of bank innovation and trade openness on market power in a multi-output banking context. In particular, we innovatively propose a causal structure analysis based on Wang and Blei (2019) to validate and verify the robustness of our results. We also assess this relationship for different types of bank ownership in China. The findings suggest that Chinese banks exhibit high market power in loans. Furthermore, the results show that bank innovation and trade openness have a significant negative impact on market power in loans, but a significant positive impact on market power in securities. The results also indicate a significantly negative impact of trade openness on overall market power. We find that higher levels of innovation among state-owned and joint-stock commercial banks improve the overall level of market power. The results suggest that, for all bank ownership types, trade openness has a significant negative impact on market power in loans but a significant positive impact on market power in securities. The impact on the overall level of market power is consistently significant and negative. / The full-text of this article will be released for public view at the end of the publisher embargo on 11 Aug 2025.

Data Envelopment Analysis (Charnes et al, 1978) is a technique used to evaluate the relative efficiency of any process or an organization. The efficiency evaluation is relative, which means it is compared with other processes or organizations. In real life situations different processes or units seldom operate in similar environments. Within a relative efficiency context, if units operating in different environments are compared, the units that operate in less desirable environments are at a disadvantage. In order to ensure that the comparison is fair within the DEA framework, a two-stage framework is presented in this thesis. Fuzzy clustering is used in the first stage to suitably group the units with similar environments. In a subsequent stage, a relative efficiency analysis is performed on these groups. By approaching the problem in this manner the influence of environmental variables on the efficiency analysis is removed. The concept of environmental dependency index is introduced in this thesis. The EDI reflects the extent to which the efficiency behavior of units is due to their environment of operation. The EDI also assists the decision maker to choose appropriate peers to guide the changes that the inefficient units need to make. A more rigorous series of steps to obtain the clustering solution is also presented in a separate chapter (chapter 5). / Master of Science