Convergence Approaches Research Articles

Towards Optimal Problem Dependent Generalization Error Bounds in Statistical Learning Theory

We study problem-dependent rates, that is, generalization errors that scale near-optimally with the variance, effective loss, or gradient norms evaluated at the “best hypothesis.” We introduce a principled framework dubbed “uniform localized convergence” and characterize sharp problem-dependent rates for central statistical learning problems. From a methodological viewpoint, our framework resolves several fundamental limitations of existing uniform convergence and localization analysis approaches. It also provides improvements and some level of unification in the study of localized complexities, one-sided uniform inequalities, and sample-based iterative algorithms. In the so-called “slow rate” regime, we provide the first (moment-penalized) estimator that achieves the optimal variance-dependent rate for general “rich” classes; we also establish an improved loss-dependent rate for standard empirical risk minimization. In the “fast rate” regime, we establish finite-sample, problem-dependent bounds that are comparable to precise asymptotics. In addition, we show that iterative algorithms such as gradient descent and first order expectation maximization can achieve optimal generalization error in several representative problems across the areas of nonconvex learning, stochastic optimization, and learning with missing data. Supplemental Material: The online appendix is available at https://doi.org/10.1287/moor.2021.0076 .

Kajian Alokasi Global Carbon Budget kepada National Carbon Budget Indonesia

Climate change and global warming are directly linked to the rise in carbon dioxide (CO2) concentration in the atmosphere, a significant byproduct of various human activities, particularly the combustion of fossil fuels. To regulate Earth's surface temperature within specified limits, the concept of a carbon budget is introduced. This budget represents the maximum cumulative carbon emissions that can be emitted to the atmosphere before the earth’s global temperature exceeds some particular temperature levels. According to the Intergovernmental Panel on Climate Change Assessment Report 6 (IPCC AR6), maintaining a temperature increase within 1,5 to 2°C requires adhering to a global carbon budget (GCB) of approximately 500 to 1.350 GtCO2 (at a 50% likelihood level). To determine the national carbon budget (NCB) for Indonesia, various allocation methods, including egalitarian, grandfathering, weighted average, and per capita convergence approaches, are applied. At a 50% likelihood level to restrict temperature rise to 1.5°C, Indonesia's NCB at the beginning of 2022 falls within the range of 14.28 to 20.55 GtCO2. These values for the carbon budget can serve as a valuable tool for assessing the climate change mitigation targets outlined in Indonesia's enhanced nationally determined contribution (E-NDC).

Revealing future trajectories in sustainable food production: convergence and openness in the innovation ecosystem

PurposeThis paper aims to explore the evolutionary dynamics of innovation ecosystems in the food industry by adopting both open innovation and convergence approaches to derive practical policy implications to develop impactful innovation ecosystems to promote food production sustainably.Design/methodology/approachStarting from the cutting-edge case of Foodvalley in the Netherlands, the study adopts a backcasting approach to reach a future vision of food ecosystems from a sustainable food production perspective. The authors set the backcasting analysis in four steps: (1) description of the present and trends analysis, (2) selection of trustworthy criteria and goals, (3) development of future images and (4) analysis of how to reach the images.FindingsThe trends analysis has identified three existing innovation fields – protein shift, circular agrifood and food and health – and two strategic directions – convergence and localness decrease. The study reports how a long-term commitment may lead the valley toward a best future scenario.Practical implicationsThe study suggests that policymakers and stakeholders can promote innovation strategies in sustainable food production ecosystems by encouraging collaboration between different sectors, reducing regulatory barriers, attracting innovative actors, and investing in education and training programs. To achieve measurable environmental and social impact outcomes, policies should promote entrepreneurship and create an enabling environment that encourages innovation and risk-taking.Originality/valueThe originality of this paper concerns the backcasting perspective applied to the analysis of the food ecosystem. This approach facilitates the identification of a path direction for successfully managing open innovation and industry convergence toward a desirable future of sustainable food production.

Towards diverging land prices in agricultural districts? Evidence from Italy before and after the great crisis

In recent decades, farmland markets have risen sharply due to their attractiveness as safe investment and savings allocation instruments. This growth has occurred globally at different rates due to the uneven expected returns, and potential land uses in different regions. This paper provides evidence of the convergence of Italian farmland markets between 1990 and 2019 by using data on average land prices with a geographically detailed resolution that considers both administrative and altitude aspects of territories. The analysis combines traditional economic convergence approaches with spatial data exploration techniques. The results are consistent and show that territorial disparities in Italian farmland markets tend to increase over the study period. This proves that divergence patterns reflect the well-established North-South dualism and disparities in socio-economic characteristics.

Graph colouring using evolutionary computation: A case study of blind naked mole‐rat algorithm

AbstractGraph colouring problem (GCP) is an NP‐complete optimization problem. It is famous for its applications in scheduling, register allocation, and map colouring. In recent years, biological inspired and especially Swarm intelligence (SI) techniques have gained popularity for solving complex optimization problems. In this article, we have proposed blind naked mole rat‐based colouring (BNMR‐Col) for graphs. BNMR‐Col uses both exploitation and exploration to find the best solution in search space. Exploitation uses both local moves and global moves to find a better solution in the surroundings of an existing solution. On the other hand, exploration generates new solution by combining various solutions from the search space. BNMR‐Col shows better convergence rate and approaches the lowest colour value in 83% of the cases when tested on standard benchmark graph instances.

At the nexus of science, engineering, and medicine: Pasteur's quadrant reconsidered.

There has been a sea change in the scientific world, advanced even more rapidly by the recent compounded public crises. Accelerated discovery, and impact from such discoveries have come from convergence approaches across disciplines, sectors, institutions, and the multiple communities seeking the common goal of innovations that transform. The classic simultaneous pursuit of fundamental understanding and application has been termed Pasteur's quadrant, where use-inspired basic research occurs. In the classic schematic developed by Donald Stokes, three quadrants represent research approaches using a 2D plane in which the vertical dimension represents the quest for understanding (basic research) and the horizontal dimension represents the consideration of use (applied research). The three outer quadrants are Bohr's (pure basic research), Edison's (pure applied research), and Pasteur's (use-inspired basic research). Viewing each of these axes as a continuum, we label the previously unnamed but contributory cell as the Innominate quadrant, where a nonzero amount of discovery and applied research also has value in generating scientific tools, novel processes or products that inform the other quadrants. More importantly, a reimagined Pasteur's quadrant schema shows a third dimension of Transformations over Time, occurring through a continuous fluid interchange among the quadrants. Transformative innovations may originate from any single quadrant. While work in Pasteur's quadrant has been shown to be highly productive, a dynamic fluid interchange among the quadrants is often involved and generates transformative advances at a faster rate. This should inform how we fund science, engineering, and medicine and educate the next generation of innovators.

Spatial patterns of regional inequalities: Empirical evidence from a large panel of countries

This paper investigates regional inequality from two standpoints. First, it explores them from a global perspective by assessing the incidence of economic growth for 2867 regions from 161 countries. Results show that middle income regions had the highest growth rates after the Great Recession, whilst regions from deciles 9 and, to a lower extent, those from decile 10 had suffered the most negative impact. Hence, the incidence of regional growth resembles the “elephant curve” of global inter-personal inequality. Second, the paper explores within-country regional inequality for a panel of 25 countries, over the period 2000–2018. Using OECD data and following both cross-sectional and time series approaches, results show that, while regional inequality decreases with growing GDPPC, regional polarization is more persistent and does not necessarily follow the same rule. The paper also delivers a first systematic assessment of sigma-convergence and regional polarization for a considerable number of countries, with findings pinpointing the importance of spatial clustering (which is contributing considerably to regional inequality levels), but also its complementarity with sigma-convergence measures (as regional inequality trajectories are not necessary associated with spatial clustering trajectories). Overall, findings also suggest that convergence approaches cannot provide an adequate theoretical background for understanding regional problems as long as they remain aspatial.

Lidar-Based Navigation of Subterranean Environments Using Bio-Inspired Wide-Field Integration of Nearness.

Navigating unknown environments is an ongoing challenge in robotics. Processing large amounts of sensor data to maintain localization, maps of the environment, and sensible paths can result in high compute loads and lower maximum vehicle speeds. This paper presents a bio-inspired algorithm for efficiently processing depth measurements to achieve fast navigation of unknown subterranean environments. Animals developed efficient sensorimotor convergence approaches, allowing for rapid processing of large numbers of spatially distributed measurements into signals relevant for different behavioral responses necessary to their survival. Using a spatial inner-product to model this sensorimotor convergence principle, environmentally relative states critical to navigation are extracted from spatially distributed depth measurements using derived weighting functions. These states are then applied as feedback to control a simulated quadrotor platform, enabling autonomous navigation in subterranean environments. The resulting outer-loop velocity controller is demonstrated in both a generalized subterranean environment, represented by an infinite cylinder, and nongeneralized environments like tunnels and caves.

Asylum burden-sharing within the EU revisited: are we moving on the right track?

ABSTRACT The refugee crisis evidenced a myriad of flaws in the EU common asylum policy, especially the lack of fairness in asylum burden-sharing. This paper reveals that, despite some progress in terms of inequality, a bi-polarised distribution gained ground. Moreover, two novel convergence approaches in this field reveal that the message of previous articles on disparity reduction is incomplete; there was no pure convergence, as it was achieved due to frontrunners, with no role for laggards. Consequently, urgent political action and greater leadership by European institutions are required to pave the way for distributive justice between states in asylum burdens.

The Implication on the Diverse Perspectives and Its Educational Applications of Architecture with Convergence Approaches in the Digital Human Era

인공지능으로 대변되는 과학기술 시대에는 인간의 창의력과 새로운 지식에의 접근, 다양한 학문을 통한 융·복합적 아이디어의 발견, 기술적 상상력과 더불어 인문학적 사고를 필요로 한다. 본 연구는 이러한 디지털 휴먼 시대에 건축을 보는 다양한 시선을 4가지 학제적(기술공학적, 도시경제환경적, 조형예술적, 인문학적) 측면으로 도출하고 이를 학교교육으로 변환하기 위한 융·복합적 접근 가능성을 모색하는 데 목적을 두었다. 기존의 기초건축교육이 국민교육을 위한 공교육으로의 도입을 시도했던 것과 달리, 현재 국토교통부가 추진하는 건축교육은 사회·경제·환경·문화적 변화에 따른 메가트렌드를 건축과 연결 하여 도시재생 및 발전에 공헌하는 건축으로의 인식 변화를 꾀하고 궁극적으로 문화적·산업적 경쟁력 향상과 건축문화 진흥의 토대 마련을 위한 교육에 집중한다. 이에 본 연구는 현재 건축 정책에서 제외된 건축이 함유한 조형예술적 의미와 인문학적 가치를 환기하여 학교교육에서의 학교급별 건축교육을 제안하면서, 디지털 휴먼 시대 인류 미래의 지속가능한 성장을 견인하는 데 있어 스마트 도시환경 정책을 실현함과 동시에 인문학적 부흥을 꾀할 수 있는 중요한 융·복합 분야가 바로 건축 이라는 점을 강조한다.

Revisiting the Solow growth model: new empirical evidence on the convergence debate

PurposeThis paper investigates income convergence using different convergence concepts and methodologies for 72 countries over the period between 1960 and 2010.Design/methodology/approachThis study applies beta (β), sigma (s), stochastic and club convergence approaches. For β-convergence analysis, it derives the cross-country growth regressions of the Solow growth model under the basic and augmented Cobb–Douglass (CD) production functions and estimates them using cross-section and panel data estimators. While it employs both the widely used coefficient of variation and recently developed weak s-convergence approaches for s-convergence, it applies three different unit root tests for stochastic convergence. To test club convergence, it estimates the log-t regression.FindingsThe results reveal that (1) there exists conditional β-convergence, meaning that poorer countries grow faster than richer countries; (2) income per worker is not (weakly) s-converging, and cross-sectional variation does not tend to fall over the years; (3) stochastic convergence is not found and (4) countries in the sample do not converge to the unique equilibrium, and there exist five distinctive convergence clubs.Research limitations/implicationsThe results clearly show that heavily relying on one of the convergence techniques might lead researchers to obtain misleading results regarding the existence of convergence. Therefore, to draw reliable inferences, the results should be checked using different convergence concepts and methodologies.Originality/valueContrary to the previous literature, which is generally restricted to testing the existence of absolute and conditional β-convergence between countries, to the best of the author’s knowledge, this is the first study to consider and compare all originally and recently developed fundamental concepts of convergence altogether. Besides, it uses the Penn World Table (PWT) 9.1 and extends the period to 2010. From this point of view, this study is believed to provide the most up-to-date empirical evidence.

CONVERGENCE OF FINANCIAL DEVELOPMENT IN ASEAN BASED ON PRIVATE CREDIT AND LIQUID LIABILITIES INDICATORS

The financial development gap for ASEAN countries is critical to analyze since the widening financial development gap will lead to underdeveloped financial sector performance. The concept considered appropriate to measure the gap of financial development is the sigma, and beta convergence approaches. Therefore, this study aims to measure, test, and analyze the convergence of sigma and beta financial development in ASEAN. The method used was the Generalized Moment Method (GMM) dynamic panel with 2010-2017. Results of the study showed that there is a convergence of sigma and beta financial development in ASEAN countries. The limitation of this study is that it only uses two indicators of financial development. The governments of each ASEAN country can improve the financial sector by reforming the financial sector to mitigate, improve, and encourage the development of the financial sector.
 Keywords: ASEAN, convergence, financial development, GMM

Equal Citizenship and Convergence

AbstractI argue against Lori Watson and Christie Hartley's recent criticisms of convergence approaches to public justification. In particular, I argue that convergence approaches can capture what is distinctive about democratic decision‐making and provide an attractive account of stability for the right reasons.

Optimistic Variants of Single-Objective Bilevel Optimization for Evolutionary Algorithms

Single-objective bilevel optimization is a specialized form of constraint optimization problems where one of the constraints is an optimization problem itself. These problems are typically non-convex and strongly NP-Hard. Recently, there has been an increased interest from the evolutionary computation community to model bilevel problems due to its applicability in real-world applications for decision-making problems. In this work, a partial nested evolutionary approach with a local heuristic search has been proposed to solve the benchmark problems and have outstanding results. This approach relies on the concept of intermarriage-crossover in search of feasible regions by exploiting information from the constraints. A new variant has also been proposed to the commonly used convergence approaches, i.e., optimistic and pessimistic. It is called an extreme optimistic approach. The experimental results demonstrate the algorithm converges differently to known optimum solutions with the optimistic variants. Optimistic approach also outperforms pessimistic approach. Comparative statistical analysis of our approach with other recently published partial to complete evolutionary approaches demonstrates very competitive results.

Assessing Financial Market Integration in Vietnam: A Price – Based Approach

This paper investigates the existence of financial market integration between Vietnam and the region (ASEAN, ASIAN) as well as the world using price-based indicators. Our paper is the first to assess the overall integration of the Vietnam financial market based on two different datasets for different sub-capital markets, the bond market and stock market, during the period of 2009 - 2018. The empirical results show that Vietnam, compared to other countries in research, holds lower regional and international integration levels in both the bond market and the stock market, but demonstrates considerably strong signs of progressing regional and global stock market integration over the period. Especially, the absolute values estimated from the beta coefficient and beta convergence approaches for the Vietnamese stock market are larger than for the bond market, proving better convergence degree and speed of Vietnamese stock market over the bond market. These bring policymakers and regulators new thoughts and more challenges in designing appropriate policies towards the development of the financial markets in the future.

Geographical Factor in Regional Development

Regional development of Russian regions is characterized by a high level of heterogeneity. Researchers highlight a significant number of factors affecting the level of regional development. Geographical factor is recognized as one of the most important factors. The development of border regions is similar to that of Russian regions. The purpose of the research is to determine the differentiation of border regions and to study the dynamics of its change. The object of the research is border regions united in the system. The process of unification of regions is based on geographical location. The subject of the study is the process of regional differentiation. Methods of assessing the convergence of economic entities are the instrument of researching regional differentiation of border regions. The convergence theorem analyzes the interregional differentiation of border regions and considers the possibility of their convergence. As a result of using convergence approaches based on the definition of convergent processes, the coefficient of variation is calculated. On the basis of the calculated values of the coefficient and its change, the possibility of convergence of border regions is assessed and the process of change of interregional differentiation is analyzed. Another direction, on the basis of which the analysis of regional heterogeneity of border regions is based, is the determination of the differentiation coefficient, which allows to determine the level of the gap between polar groups of regions. The goal set in the study will allow assessing the impact of geographical location on economic development and formulating directions of economic policy. Spatial heterogeneity of border regions defined by means of mathematical tools can be used when writing the Program and the Strategy for Regional Development, within which one of the competitive advantages will be the location of the territory.

Machine Learning Algorithms for MR Brian Image Classification

Effects of Visual and Auditory Response Times in Males and Females by using Tactile and Mouse on Web Based EnvironmentClassification of MR brain images accurately is the crucial part in the medical imaging research since brain is a complex structure which needs multiple convergence approaches to classify. In this paper machine learning algorithms like SVM, HMM were employed to classify the brain images with the use of spatial features. Super pixel image segmentation approach is applied to segment the image into distinct super pixel region for which the features are calculated.

The vector of state government participation in the innovation process: federal and regional aspects

Introduction. The article presents the authors' concept of state government as a regulator of the innovation process. The transition to an ongoing innovation process is an important feature of the modern period of scientific and technological development of the world economy. However it is impossible to approach the scientific and technological level required for developed countries in a centralized regulatory manner using only direct financing without the formation of innovative infrastructure. Materials and methods . The authors used the data of the Russian Statistic Bulletin and the analytical materials the Far East Investment and Exports Support Agency. The methodological basis includes the convergence and holistic scientific methods and approaches inherent in the institutional-evolutionary direction. Research result . As a result of the analysis, a new vision of the state government role as a regulator which is necessary to achieve the goals of scientific and technological development in the context of commercialization of the innovation process is outlined. Reducing the public functions does not contradict the concept of the active position of the state government in the formation of the institutional and infrastructure environment. The authors substantiate that it is necessary at the stage of forecasting and formation of the regional strategy. This point follows from the main directions of the economic policy on improving the investment climate in the scientific and technological sphere of the Far Eastern Federal District and the experience of the countries of the Asia-Pacific region, which the Russian Far East belongs to. The hierarchy of structures of innovative development is offered: The Federal Executive authorities - the institutional level represented by the Institute coordinating investment development and the financial instrument of development - administrative services of the entity of the Russian Federation - public institutions. Together these structures form an organizational mechanism for the development and implementation of the national innovation system development strategy. The formation of investment institutions complements the hierarchy of innovation development structures. Regional institutions created to promote the improvement of the investment climate should at the same time be innovation development institutions built into the hierarchy of policy levels of the regulator. Discussion and conclusion . The article discusses the evaluation of possibilities and conditions for combining and distribution by the Federal and Regional Authorities of functions to organize business engagement in the scientific-technological sphere and to overcome the regional and sectoral barriers. The authors substantiate the principles of a new technical and economic paradigm of innovative development: reducing the public functions, changing the nature of public institutions, changing the financial architecture, updating and mobility of production factors, virtualization of society.

The Convergence of Food Expenditures in the European Union Countries – a Spatio‑Temporal Approach

The paper presents the analysis of the convergence of household final consumption expenditures on food in the European Union countries in the period of 1999–2015, considering spatial and spatio‑temporal tendencies and dependencies. The aim of this research is to verify whether space significantly influences the convergence of the considered process. The subject of the investigation is the share of household final consumption expenditures on food in total final consumption expenditures of European countries. Spatial and spatio‑temporal tendencies and dependencies are surveyed using the conception of spatial and spatio‑temporal trends and spatial autocorrelation. The convergence of the process is investigated with the use of β‑convergence models for panel data (also in the spatial terms). Absolute and conditional convergence approaches are applied. For the conditional approach, the models are expanded to incorporate the influence of additional determinants, including space, disposable income and the level of food prices.

Coupling a ladder to the safety net: reinventing MGNREGA for asset creation

ABSTRACTIndia’s rural employment act (MGNREGA), the world’s largest public works programme, is designed to guarantee wage employment as a statutory right. It is viewed by many as a drain on the public exchequer, as wages are paid but durable assets are not created. Can this safety net be restructured to double up as a ladder of opportunity for the poor without altering its entitlement-based framework? This article shows that implementing four sets of interventions – policy for household livelihood assets, convergence approaches, pro-poor participatory planning, and strengthening accountability, can accelerate the creation of livelihood assets in the lands of the poor.