Data Envelopment Analysis Efficiency Research Articles

Performance Measurement: Machine Learning as a Complement to DEA for Continuous Efficiency Estimation

Data Envelopment Analysis (DEA) is a well-established non-parametric technique for performance measurement to access the efficiency of Decision-Making Units (DMUs). However, its inability to predict the efficiency values of new DMUs without re-conducting the analysis on the entire dataset has led to the integration of Machine Learning (ML) in previous studies to address this limitation. Yet, such integration often lacks a thorough evaluation of ML's adaptability in replacing current DEA process. This paper presents the results of an empirical study that employed eight ML models, two DEA variants, and a dataset of S&P500 companies. The findings demonstrated ML’s remarkable precision in predicting efficiency values derived from a single DEA run and comparable performance in predicting the efficiency of new DMUs, thus eliminating the need for repeated DEA. This discovery highlights ML’s robustness as a complementary tool for DEA in continuous efficiency estimation, rendering the practice of re-conducting DEA unnecessary. Notably, boosting models within the Ensemble Learning category consistently outperformed other models, highlighting their effectiveness in the context of DEA efficiency prediction. Particularly, CatBoost demonstrated its superiority as the top-performing model, followed by LightGBM in the second position in most cases. When extended to five enlarged datasets, it shows that the model exhibits superior R² values in the CRS scenario.

Multicast Routing Based on Data Envelopment Analysis and Markovian Decision Processes for Multimodal Transportation

In the context of Iraq’s evolving transportation landscape and the strategic implications of the Belt and Road Initiative, this study pioneers a comprehensive framework for optimizing multimodal transportation systems. The study implemented a decision-making framework for multimodal transportation, combining data envelopment analysis (DEA) efficiency scores and a Markov decision process (MDP) to optimize transportation strategies. The DEA scores captured decision-making unit (DMU) performance in various aspects, while the MDP rewards facilitated strategic mode selection, promoting efficiency, cost-effectiveness, and environmental considerations. Although our method incurs a total cost approximately 29% higher than MRMQoS, it delivers a nearly 26% reduction in delay compared to MCSTM. Despite MRMQoS yielding an 8.3% higher profit than our approach, our proposed scheme exhibits an 11.7% higher profit compared to MCSTM. In terms of computational time, our method achieves an average CPU time positioned between MCSTM and MRMQoS, with MCSTM showing about 1.6% better CPU time than our approach, while our method displays a 9.5% improvement in computational time compared to MRMQoS. Additionally, concerning CO2 emissions, the proposed model consistently outperforms other models across various network sizes. The percentage decrease in CO2 emissions achieved by the proposed model is 7.26% and 31.25% when compared against MRMQoS and MCSTM for a network size of 25, respectively.

An innovative environmental regulation tool for regional air pollution based on DEA-RFR

Environmental regulation tools are one of the conventional means for pollution control in various countries. For large-scale urbanization regions, reasonable pre-allocation of pollutant emission amount for each area is an effective regulation to improve air quality. Treating pollutant emission quota as a resource element, a comprehensive input-output model has been established based on Data Envelopment Analysis (DEA) method, to evaluate the rationality of pollutant emission allocation and provide adjustment scheme, thus developing an environmental regulation tool. The index system includes factors of economy, environment, energy consumption, demographic and land-use as input indicators, regional GDP and environmental air quality index as output indicators. The regulation tool is applied to the current air pollutant emission management in Henan Province, China. The results show that 7 cities need to further adjust the allocation scheme, among which, Anyang, Xinxiang, Nanyang and Shangqiu are suggested to reduce the emission allocation quotas; Puyang and Zhumadian are suggested to increase the emission allocation quotas; Pingdingshan shall seek optimization paths through collaborative improvement. Further, to verify the effectiveness of the tool, by joint using Random Forest Regression (RFR), a DEA efficiency reassessment is conducted on the obtained new allocation scheme of Nanyang City. The results show that the efficiency of the new allocation scheme has reached optimal. This study provides a practical and effective regulation tool for regional management of pollutant emission, to balance economic growth and environmental protection.

Evaluation of the efficiency of agricultural higher education institutions of Ukraine and the dynamics of its change by data envelopment analysis

Presently, the process of assessing the effectiveness of higher education institutions and analysing the dynamics of its change needs to be improved, which determines the relevance of this study. The purpose of this study was to improve the assessment of the efficiency of agricultural higher education institutions of Ukraine in the scientific and journalistic context, as well as to investigate the changes in dynamics using the non-parametric method of assessing the relative efficiency of data envelopment analysis. The study used methods of statistical processing and data analysis, linear programming, computer computing to solve research problems, and logical and analytical methods to determine the purpose of the study, the means of achieving it, substantiate the findings of the study, and formulate conclusions. To evaluate the effectiveness of scientific and publishing activities of these higher education institutions, the input-oriented model of variable returns to scale of the data envelopment analysis method was applied. Using statistical data for 2020- 2022, the net technical efficiency of scientific and publishing activities of 37 agricultural higher education institutions of Ukraine was assessed. Based on the results of the assessment, a rating of agricultural higher education institutions in Ukraine was compiled. It was found that in 2021 and 2022, 27% of the surveyed higher education institutions were 100% effective. The average net technical efficiency in 2021 and 2022 was 0.91. In 2021-2022, 27% of agricultural higher education institutions increased the efficiency of scientific and publishing activities, 59% did not change, and 14% decreased. On average, there was a 1% increase in the use of educational resources in the field of agricultural higher education. Thus, the scientific and publication activities of agricultural higher education institutions during the study period are progressive. The findings of this study can be used for a reasonable ranking of higher education institutions according to the efficiency of scientific and publishing activities, analysis of the dynamics of changes in efficiency.

Financial efficiency and financial resource allocation of Beijing–Tianjin–Hebei urban agglomeration

This study focuses on improving financial efficiency and rationally allocating financial resources in the Beijing–Tianjin–Hebei, based on data from 2011 to 2019. Firstly, the Slack-Based Measure (SBM) model based on Shannon Entropy is adopted to measure financial efficiency in Beijing–Tianjin–Hebei. Secondly, a Beijing–Tianjin–Hebei inter-regional financial resource allocation scheme is proposed using the Generalized Equilibrium Efficient Frontier Data Envelopment Analysis (GEEFDEA) model. Finally, the external environmental factors influencing financial efficiency are explored using spatial and other econometric models. Efficiency measurement reveals that financial efficiency in Beijing–Tianjin–Hebei is unevenly developed. The financial efficiency of Beijing, Tianjin, and Zhangjiakou is high and stable, whereas the financial efficiency of Tangshan, Langfang, and Shijiazhuang is high, but fluctuates significantly. In comparison, the financial efficiency of other cities is low. The factors and the adjustment amount to improve urban financial efficiency are obtained by studying the allocation of financial resources. And it is found that Tianjin and Tangshan have more financial resource surpluses than other cities. By exploring influencing factors, it is found that the financial development level, innovation level, and infrastructure construction level of Beijing–Tianjin–Hebei significantly affect financial efficiency.

Government spending efficiency in Latin America

We assess public spending efficiency of 20 Latin American countries over the period of 2000–2019, computing data envelopment analysis efficiency scores. For the Public Sector Performance composite indicator, we use the annual data of socio-economic indicators, and for the input measure we consider Total Public Spending as a percentage of GDP, by spending category. The results show that public spending during the period under study increased, but that overall governments were not efficient, as on average they could have used 27% less spending to achieve the same levels of performance. On the other hand, governments could have increased their performance by 18% whilst maintaining the same level of spending. The most-efficient countries were Chile, Guatemala, Panama, and Paraguay, with the least efficient being Bolivia, Venezuela, Nicaragua, Suriname, and Brazil.

Research on the Operational Efficiency of Basic Medical Insurance for Urban Employees in China-Based on Three-Stage DEA Model.

To conduct a three-stage data envelopment analysis (DEA) of the operational efficiency of urban employee basic medical insurance (UEBMI) across 31 provinces in China from 2012 to 2021. Utilized a three-stage DEA model, this study measured and evaluated the operational efficiency of urban employee basic medical insurance within China's 31 provinces. The operational efficiency of urban employee basic medical insurance in China was notably low, displaying significant disparities across different regions and periods. Substantial room for improvement exists. Environmental factors, including urbanization level and unemployment rate, wielded a pronounced influence on the operational efficiency of China's employee medical insurance. Conversely, the fiscal revenue and expenditure ratio hindered the enhancement of employee medical insurance efficiency. This study held valuable insights for enhancing the operational efficiency of China's urban employee basic medical insurance. To effectively improve insurance efficiency, the following recommendations are put forth: firstly, relevant departments should escalate resource investments and optimize resource utilization based on rational allocation; secondly, enhance legislation and regulations, fortify fund oversight, and ensure equitable and judicious utilization of the medical insurance fund; finally, leveraging modern high-tech advancements can comprehensively elevate the operational efficiency of the employee medical insurance fund. Ultimately, with the advent of cutting-edge technology, the operational efficiency of employee medical insurance fund can be comprehensively elevated.

An Exploration of Prediction Performance Based on Projection Pursuit Regression in Conjunction with Data Envelopment Analysis: A Comparison with Artificial Neural Networks and Support Vector Regression

Data envelopment analysis (DEA) is a leading approach in performance analysis and discovering newer benchmarks, and the traditional DEA models cannot forecast the future efficiency of decision-making units (DMUs). Machine learning, such as the artificial neural networks (ANNs), support vector machine/regression (SVM/SVR), projection pursuit regression (PPR), etc., have been viewed as beneficial for managers in predicting system behaviors. PPR is especially suitable for small and non-normal distribution samples, the usual cases in DEA analysis. This paper integrates DEA and PPR to cover the shortcomings we faced while using DEA and DEA-BPNN, DEA-SVR, etc. This study explores the advantages of combining these complementary methods into an integrated performance measurement and prediction model. Firstly, the DEA approach is used to evaluate and rank the efficiency of DMUs. Secondly, we establish two DEA-PPR combined models to describe the DEA efficiency scores (also called the production function) and the DEA-efficient frontier function. The first combined model’s input variables are input–output indicators in the DEA model, and the output variable is the DEA efficiency. In the second model, its input variables are input or output indicators in the DEA model, and the output variable is the optimal input indicator for input-oriented DEA or the output indicator for output-oriented DEA. We conducted positive research on two examples with actual data and virtual small, medium-sized, and large samples. Compared with the DEA-BPNN and DEA-SVR models, the results show that the DEA-PPR combined model has more vital global optimization ability, better convergence, higher accuracy, and a simple topology. The DEA-PPR model can obtain robust results for both small and large cases. The DEA-BPNN and DEA-SVR models cannot obtain robust results for small and medium-sized samples due to overfitting. For large samples, the DEA-PPR model outperforms DEA-BPNN, DEA-SVR, etc. The DEA-PPR combined model possesses better suitability, applicability, and reliability than the DEA-BPNN model, the DEA-SVR model, etc.

Smart Heating, Ventilating, Air-conditioning and Refrigeration by Web-based Geographic Information System

Objective: Heating, ventilating, air-conditioning and refrigeration (HVAC&R) systems currently account for a significant portion of energy consumption. The HVAC&R system contributes the largest energy consumption in a building, so it is essential to optimize energy consumption to improve energy saving worldwide. Methods: The web-based geographic information system (GIS) enables the seamless sharing of spatial data across the globe, accessible anytime and anywhere via the World Wide Web. The set of remote reading networked sensors, advanced modems, and data loggers facilitate the intercommunication for the geodatabase of HVAC&R’s facilities. The integration of remote sensing technology and the Internet of Things, grounded in GIS establishes a control loop dedicated to energy conservation. This method is a pioneering concept in control science, offering significant potential for enhancing design, maintenance, and energy management practices. It empowers energy users with real-time control over their energy consumption, making a substantial advancement in this field. Results: In this work, the model of HVAC&R control in context with web-based GIS showed that the regression mathematical analysis in compliance with the computational method holds the capacity to predict energy consumption and evaluate energy loss. Conclusion: In regression analysis, the P was found to be 0.991 for the percentage of dissatisfaction, 0.977 for energy use intensity, and 0.962 for data envelopment analysis efficiency. Additionally, the curve estimation showed that the power function was utilized in regression analysis processes.

Data Envelopment Analysis and hyperbolic efficiency measures: Extending applications and possibilities for between-group comparisons

Data Envelopment Analysis (DEA) is a widely used tool to estimate relative efficiencies. However, this methodology encounters issues when between-group comparisons with variable returns of scale (VRS) are of interest. In this specification efficiency scores can be undefined when the projection of the efficiency estimation of an observation from one group does not have an intersection with the frontier of the other group. To address this issue, a hyperbolic efficiency estimation can be used. Up to now, the literature lacks consideration of weight restrictions and non-discretionary variables in the hyperbolic DEA method. Weight restrictions allow redefining the production possibility set (PPS), while non-discretionary variables adjust the model’s orientation. Considering these factors, it is possible to incorporate prior knowledge, substitution effects, and increase discriminatory power. This paper introduces the respective mathematical formulations and the practical implementation in statistical software. The estimation methods have been combined in an R-package called hyperbolicDEA in order to facilitate the application of hyperbolic DEA with weight restrictions, non-discretionary variables, and additional functionalities. An empirical example of fish farms illustrates the advantages of the introduced methodologies and the functionalities of the R-package.

Maximum slacks-based measure of efficiency in network data envelopment analysis: A case of garment manufacturing

The conventional slacks-based measure (SBM) model is widely applied to measure the relative efficiency of a set of decision making units (DMUs) because the measured efficiency is not affected by weakly efficient frontiers. This model seeks the target point on the production frontier that is farthest to the assessed DMU to calculate efficiency, which requires more effort for the DMU to reach to become efficient. This paper develops a model to calculate the maximum SBM efficiency for general network production systems. Different from the conventional minimum SBM efficiency that is always less than the radial efficiency for inefficient DMUs, the maximum SBM efficiency can be greater than the radial efficiency. The measured efficiency satisfies the monotonicity property. The efficiencies of the component divisions of the network system can also be calculated at the same time. A garment manufacturing company whose major operations comprise cutting and sewing is studied using the developed model. The results show that the average maximum SBM efficiency is 35 % higher than the average minimum SBM efficiency and the sewing operation contributes more to the system efficiency than the cutting operation does.

A study on the possibility of a BUS route network with connections based on an analysis of the efficiency of the BUS route network and resistance OF BUS users to transfer

In this study, data envelopment analysis (DEA) was used to evaluate the efficiency of the bus route network in Akita City, Akita Prefecture, Japan. The DEA efficiency indices utilized route length (km) and overlap index (number of lines/km) as input values for the network efficiency, while the number of buses per year was used as the output value for the deployment efficiency. Operating costs (yen) were used as the input value for the revenue efficiency, and Operating costs (yen) and fare revenue (yen) were used as input and output values, respectively, for the revenue efficiency. The results of this study show that the overall network efficiency is low, indicating that complex and radial bus routes are inefficient and can be improved. Next, we surveyed the attitudes of residents of Akita City, regarding the use of bus routes. The results showed that 18.5% of the respondents did not want to transfer between bus trips, whereas 69.7% could tolerate transferring up to once. Furthermore, 2.7% of the respondents were unwilling to wait for the bus, whereas 52.4% were willing to wait within 10 min, while 38.6% were willing to wait between 10 and 20 min.Based on the results of this analysis, which was conducted by a university laboratory, suggestions were made to the local government on how to provide an efficient network. The bus routes in Akita City are operated by private operators but managed by the local government. Thus, efficient bus routes with connections can be constructed through proposals to local governments.Inefficient bus routes were identified using DEA, and a network with transfers was examined for their improvement. When examining such a network, the attitudes of the users toward transfers must be understood; the relationship involves understanding the attitudes of the users toward inefficient bus routes using DEA. This study revealed that a large number of people could tolerate up to one transfer. The transit time is a significant factor in the acceptance.

Assessment of Interorganizational Technology Transfer Efficiency

In this paper we propose a methodology for assessing the efficiency of technology transfer through merger and acquisition (M&A) and empirically estimate the effect of key factors impacting it. We implement data envelopment analysis (DEA) to calculate an efficiency score of the technology transfer process. The DEA efficiency score integrates a set of outputs (post-merger characteristics of an acquirer) and inputs (pre-merger technological parameters of a target); thus, it provides a multidimensional estimate of efficiency adjusted for the value of the acquired technology base.In the empirical part of this research, we collect data from 434 M&As to study a channel for transferring a technology base across organizational boundaries. Overall, empirical results suggest the adverse outcomes of accumulation of capability to value external technology: the higher the acquirer’s R&D intensity, the lower the efficiency of interorganizational technology transfer. The size of acquirer and relative size of the deal also affect the post-merger outcomes significantly and negatively. At the same time, the estimated effect of such technological characteristics of acquirer as capital expenditure intensity and number of patents is insignificant.

Quantifying the impact of COVID-19 on Chinese ports

PurposeThe purpose of this study is to quantify port efficiency assessment indicators to analyze the impact of COVID-19 on Chinese One Belt One Road (OBOR) ports.Design/methodology/approachThis study utilized a grey prediction model GM(1,1) to forecast five relevant indicators for each of the 17 OBOR ports both with and without COVID-19 background conditions. Additionally, the data envelopment analysis (DEA) efficiency assessment approach was used to analyze the impact of COVID-19 on port efficiency.FindingsThe results indicate that cargo and container throughput growth rates during the COVID-19 pandemic are reduced by 1.7 and 2.1%, respectively. There was also a noticeable reduction in technological efficiency (TE) as well as pure technological efficiency (PTE), while scale efficiency (SE) remained largely unaffected. Furthermore, the dynamic efficiency MI was mainly negatively impacted by changes in overall efficiency change (EFFCH), where pure efficiency change (PECH) less than one contributed significantly towards overall regression of port efficiencies during this period.Originality/valueThis paper is unique in its use of a combination of the grey prediction model and DEA efficiency assessment to quantify changes in important indicators during pandemic periods. This approach not only provides a quantitative understanding of the impact on port-level efficiency through numerical quantification but also offers readers an intuitive understanding.

CO2 Emissions in Air Transport: A Comparative Study Using Data Envelopment Analysis

The international community has been actively addressing the issue of reducing carbon dioxide (CO2) emissions, as evidenced by various initiatives such as Rio 92, Rio +20, the Paris Agreement, and the 2030 Agenda. These endeavors aim to promote sustainability worldwide. While air transport accounts for only 2% of anthropogenic CO2 emissions, its projected growth has spurred the exploration of new technologies to enhance aircraft efficiency. Consequently, further studies are necessary to assess the efficiency and impact of air transport on CO2 emissions in most countries. A methodology utilizing data envelopment analysis (DEA) is proposed in this study for this purpose. This methodology correlates the structure and annual air movement of 21 countries with CO2 production, considering it an undesirable output. The clustering method establishes the relationship between DEA efficiency, the number of airports, air traffic movements (landings and take-offs), and annual economic factors such as gross domestic product in these countries. Additionally, a regression analysis is conducted to examine the influence of air transport on CO2 emissions from transport systems. The findings reveal that the U.S. and China exhibit low efficiency as outliers because of their significant economic disparities. Other countries are grouped based on economic similarities, with South American nations such as Colombia and Chile showing lower air traffic movement efficiency compared with economically powerful European countries such as the UK and Germany. However, the impact of air transportation variables on CO2 emissions remains substantial.

Microfinance Institutions’ Efficiency and its Determinants: Evidence from India

This study evaluates the financial and social efficiency of Indian Microfinance Institutions (MFIs) from 2005 to 2018 and also tries to find out the determinants of financial and social efficiency. In the first step, bias-corrected bootstrap Data Envelopment Analysis (DEA) efficiency scores for financial and social efficiency were calculated using two input and two output variables. In the second step of analysis above efficiency scores have been used as dependent variables and Seemingly Unrelated Regression (SUR) model is employed to ascertain the determinants of financial and social efficiency of Indian MFIs. The study finds that the aggregate financial efficiency of Indian MFIs is higher than the social efficiency. Over the study period, efficiency of Indian MFIs shows an increasing trend. Results of SUR suggest that as compared to Non-Banking Finance Company (NBFC) MFIs, Non-NBFC MFIs’ financial and social efficiency is better. Influence of age, scale of loan portfolio, asset quality and financial leverage on efficiency of MFIs are insignificant. This study also finds that asset size of the firm is positively associated with efficiency. Evidence from this study suggests that reform measures taken post microfinance crises has negatively affected the financial efficiency. However, social efficiency has improved during post reform period. JEL Codes: G21, C33

Cross efficiency model of network DEA and its application on low carbon efficiency evaluation of multimodal transport

The transportation industry is the main source of carbon dioxide (CO2) emissions, and the development of multimodal transport is conducive to reducing transportation carbon emissions. In order to evaluate the low carbon efficiency of multimodal transport, this article proposes a cross efficiency data envelopment analysis (DEA) method for complex network structures from the perspective of fairness, and takes China's rail-water intermodal transport as an example for research. The proposed method includes two secondary goal models that reduce the unreasonable differences between self-evaluation and peer-evaluation and treat all stages within the network system equally. Then, the improved entropy weight method is used to aggregate the cross efficiency. In addition, a numerical example is given to verify the effectiveness of the proposed method. The results of the case study show that the efficiency of Tangshan Port and Rizhao Port is relatively high, and others should learn from the successful experiences and develop improvement strategies based on their own characteristics. The conclusions can provide reference for transportation enterprise management and policy formulation, and the proposed method can be widely applied to the performance analysis of enterprise network, industrial chain, and supply chain.

Dealing the problem of heteroscedasticity in Data Envelopment Analysis

There are various performance or efficiency measurements of similar type of organizations known as Decision Making Units (DMU). One of the effective tools for assessing the effectiveness of DMUs is the non-parametric Data Envelopment Analysis (DEA) approach. However, the presence of heteroscedasticity in the inputs and outputs affect the results of DEA. By computing the efficiency estimates in this work, the author tried to eliminate the heteroscedasticity. Additionally, for the sake of an empirical inquiry, the author took into account information on non-life insurance businesses in India. The study's goal is to determine how heteroscedasticity affects DEA efficiency estimations.

Data Envelopment Analysis for the Selection of Highway–Rail Grade Separation Projects

With the ambitious goal of zero deaths but growing traffic and congestion at highway–rail grade crossings (HRGCs), grade separation of these conflicting points has become an increasing strategy that improves railway safety but also increases network capacity and enhances livability attributes. Meanwhile, the engineering costs of grade separation are expensive and can vary considerably, which results in difficulties in the rational selection of such projects. This paper implemented the data envelopment analysis (DEA) for the benchmarking of high-risk HRGCs for grade separation considerations in Oklahoma. The top 1% of HRGCs were identified based on the New Hampshire Hazard Index (NHI) rankings and used as the decision-making units (DMUs) for DEA. The outputs of the DEA process included safety benefits, travel time savings, vehicle operating cost savings, and environmental benefits, and the input was the estimated cost of a grade separation project. The DEA efficiency scores of the DMUs were calculated, and nine HRGCs were identified on the frontiers with efficiency scores of 1.0. To further discriminate the DEA-efficient HRGCs, the enhanced DEA (E-DEA), which combines self-evaluation and cross-evaluation methods, was performed and the most efficient crossing was decreased to four with the E-DEA scores of 1.0. This study established a hierarchical framework combining the NHI, DEA, and E-DEA approaches, available data sets, and decision-making tools for the selection of the most efficient investment alternatives for highway–rail grade separation.

Technical efficiency evaluation and ranking of container terminal operators through DEA

In today’s competitive business environment the evaluation and ranking of container terminals plays a crucial role. Data Envelopment Analysis (DEA) models are suitable to measure the efficiencies of alternatives specified with multiple inputs with multiple outputs, where as Stochastic Frontier Analysis (SFA) models are suitable to measure the efficiencies of alternatives specified by multiple inputs with single output. In this four inputs and four outputs are considered for the evaluation of technical efficiency of container terminals and DEA is implemented to evaluate the performance of these container terminals based on technical efficiency. Also Regression analysis is performed to identify the factors which affect the technical efficiency mostly.