Structure Of Innovation Networks Research Articles

Workload prediction based on improved error correlation logistic regression algorithm and Cross‐TRCN of spatiotemporal neural network

AbstractIn view of the randomness of user network usage behavior in data centers, which leads to a large randomness in power load, and considering that a single randomness processing method is usually difficult to fully characterize the uncertain characteristics of the system, this paper proposes a dual fusion prediction analysis model based on an improved error correlation logic regression algorithm and a novel spatiotemporal neural network structure called Cross‐TRCN. Two weight coefficients λ1 and λ2 are introduced to fuse the prediction results with different long‐term sequence prediction performance, thereby further eliminating the influence of random errors. The results show that it is feasible to predict the workload of data centers based on the improved error correlation logic regression algorithm and the innovative spatiotemporal neural network structure Cross‐TRCN.

Drivers for Inter-city Innovation Networks Across Chinese Cities: Modelling Physical Versus Intangible Effects

AbstractCross-region innovation is widely recognized as an important source of the long-term regional innovation capacity. In the recent past, a growing number of studies has investigated the network structure and mechanisms of cross-region innovation collaboration in various contexts. However, existing research mainly focuses on physical effects, such as geographical distance and high-speed railway connections. These studies ignore the intangible drivers in a changing environment, the more digitalized economy and the increasingly solidified innovation network structure. Thus, the focus of this study is on estimating determinants of innovation networks, especially on intangible drivers, which have been largely neglected so far. Using city-level data of Chinese patents (excluding Hong Kong, Macao, and Taiwan Province of China), we trace innovation networks across Chinese cities over a long period of time. By integrating a measure on Information and Communications Technology (ICT) development gap and network structural effects into the general proximity framework, this paper explores the changing mechanisms of Chinese innovation networks from a new perspective. The results show that the structure of cross-region innovation networks has changed in China. As mechanisms behind this development, the results confirm the increasingly important role of intangible drivers in Chinese inter-city innovation collaboration when controlling for effects of physical proximity, such as geographical distance. Since digitalization and coordinated development are the mainstream trends in China and other developing countries, these countries’ inter-city innovation collaboration patterns will witness dramatic changes under the influence of intangible drivers.

Evolution game model and simulation of enterprise R&D cooperative innovation network under network externalities

ABSTRACT Multi-agent cooperation, pooling innovative resources, helps to develop key technologies and major projects in enterprises. However, the multi-agent cooperation of enterprises makes the cooperative relationship unstable, and behaviours such as moral hazard, spillover of technology or information are more likely to occur, thereby increasing the risk of enterprises. In this research an enterprise R&D cooperative evolutionary game model is built to investigate the influence of Network externalities on enterprise R&D cooperation in innovation networks. Furthermore, the evolution of innovative network structure of game results is simulated by means of computer technology so that the effects of network externalities on corporation degree, agglomeration and network structure are analyzed. The results show that Network externalities led to a decrease of cooperated partners, a decline in cooperation degree and agglomeration of enterprises, and a shrinking enterprise R&D cooperation tendency. The influence of network externalities shows significant structural differences in the three types of enterprise networks: the degree of enterprise cooperation and agglomeration fluctuates greatly in the scale-free network, but slightly in the small-world network and the random network.

Spatial pattern characteristics and optimization policies of low-carbon innovation levels in the urban agglomerations in the Yellow River Basin

This paper refines a comprehensive index of urban low-carbon innovation based on urban low-carbon innovation concept and an urban low-carbon innovation community structure and function. This index is complemented by an indicator reflecting equity in terms of urban-rural income gap, some indicators of low-carbon knowledge and technology innovation, as well as an indicator of the characteristics of a carbon circular economy, and a further refinement of the indicator of low-carbon innovation system. We use the entropy value method, kernel density analysis, gravity model, and social network analysis to study the low-carbon innovation spatial pattern in urban agglomerations (UAs) of the Yellow River Basin (YRB). Innovations in the findings are: (1) the low-carbon innovation level in the UAs in the YRB decreases in spatial distribution from the downstream to the midstream to the upstream UAs. (2) Reveal the spatial pattern laws of low-carbon innovation levels in different types of UAs in the YRB. The spatial pattern in regional level UAs is more complex and hierarchical than that in area level UAs; but there are still gaps in the hierarchy of spatial patterns of low-carbon innovation levels between UAs at the regional level, as well as between UAs at the area level. The spatial pattern of two regional-level UAs in the YRB is transitioning from a low-carbon innovation point-axis to a low-carbon innovation network structure, and the spatial pattern of one regional-level UAs is at the embryonic stage of the low-carbon innovation point-axis structure. The spatial pattern of one area-level UA is at the embryonic stage of a low-carbon innovation point-axis structure, and the spatial pattern of three area-level UAs is at the stage of low-carbon innovation pole. The spatial pattern of low-carbon innovation becomes more complex and hierarchical as we move from upstream to midstream to downstream UAs. (3) Use social network analysis and the study show that the low-carbon innovation network is dominated by the internal correlations among the UAs in the upstream, midstream, and downstream. The low-carbon innovation linkages between upstream, midstream and downstream UAs are weak. Finally, we put forward corresponding policy recommendations.

THE INFLUENCE OF FIRM KNOWLEDGE CHARACTERISTICS ON TECHNOLOGICAL INNOVATION: A MULTILEVEL NETWORK STRUCTURE PERSPECTIVE

This study describes collaboration between firms and different organizations as a process of knowledge sharing, integrating inter-organizational collaboration networks and knowledge networks into a unified framework. It analyzes the bilayer structure of collaborative innovation networks, namely, the firm-knowledge bipartite network. The study investigates the impact of firm knowledge characteristics (diversity and uniqueness) on breakthrough and incremental innovation within collaborative innovation networks. The research employs patent data from Chinese A-share high-tech manufacturing companies listed from 2000 to 2018 and verifies the related hypotheses using a negative binomial regression model. The results suggest that, within collaborative innovation networks, firm knowledge diversity has a negative impact on breakthrough innovation but a positive impact on incremental innovation. On the other hand, firm knowledge uniqueness plays a positive role in breakthrough innovation but a negative role in incremental innovation. This study extends the understanding of firm technological innovation by considering both breakthrough and incremental innovation as distinct behaviors, providing a new perspective on the mechanisms underlying different innovation behaviors. By adopting a multilevel network structure approach to collaborative innovation, it contributes to a deeper understanding of internal and external factors influencing firm innovation behavior and expands the application scope of social network analysis.

Optimization of school-enterprise cooperation path for engineering majors based on robust optimization algorithm

Abstract This research delineates the unique attributes of the industry-university-research collaboration network, formulating an innovation network specifically designed for engineering disciplines predicated on university-industry partnerships. It introduces a set of measurement indices aimed at evaluating the structural properties of this network. To address the prevalent dissatisfaction within conventional collaborative network decision-making scenarios, this study proposes a theoretical framework comprised of “baseline,” “target,” and “status quo” strategies. This framework systematically standardizes attribute values and constructs a decision-maker satisfaction function. Utilizing this function, the study identifies the optimal solution by ranking the satisfaction levels associated with various alternatives within the cooperative innovation network. Further, this paper examines the characteristics and evolutionary trends of the regional network structure of industry-university-research cooperation in the context of 5G technology, encompassing engineering departments, educational institutions, and corporate entities. It also analyzes the optimization outcomes of six preferred linkage types within the engineering majors’ university-enterprise industry-university-research innovation network under a scenario of enhanced network relationships. By comparing the weighted degree distributions between the optimally optimized network and the actual network and employing the entropy value method along with the TOPSIS method, the study evaluates the disparities in network structures between the two scenarios and identifies the optimal connection paths. The comprehensive score of the optimization effect of university-enterprise-industry-university-research cooperation and innovation network structure of engineering majors in the context of network relationship strengthening is ranked as Q 3 > Q 5 > Q 6 > Q 2 > Q 4 > Q 1, and the comprehensive score of the strong-weak connection type is the highest at 0.6774. That is, maintaining the connecting edges between “strong enterprise” nodes and “weak enterprise” nodes in the industry-university-research cooperation and innovation network has the best effect on the optimization of university-enterprise industry-university-research cooperation and innovation network for engineering majors.

The role of Horizon (2020) in achieving climate-neutral urban objectives: A study of 14 Spanish cities innovation networks

This study aims to analyse the structure and characteristics of innovation networks from Horizon (2020) projects in fourteen Spanish cities, and to understand how these networks influence a city's success in being selected for the European Mission on Climate-neutral and smart cities by 2030. Utilizing social network analysis, the research uncovers patterns within these urban innovation networks, emphasizing attributes that advance climate neutrality efforts. Results indicate that cities with larger and more interconnected innovation networks, exhibiting balanced participation from diverse stakeholders, and predominantly led by the private sector and research institutions, are more likely to be successfully chosen for the Cities European Mission. This research provides a novel insight into the role of urban innovation systems in the journey towards decarbonization, accentuating the significant influence of European institutional structures in moulding these networks. By focusing on urban networks as a metric for assessing the progress towards climate neutrality, this study offers a fresh perspective that diverges from traditional regional or transnational analyses. The findings have substantial implications for policymakers and urban planners, underscoring the need for fostering robust innovation networks as a strategy to achieve climate neutrality. Additionally, the results pave the way for future research on optimizing resource allocation within urban innovation systems, ultimately contributing to the broader objectives of the European Green Deal.

The Study of Regional Innovation Network Structure: Evidence from the Yangtze River Delta Urban Agglomeration

As a driving force for regional development, innovation holds an increasing position in regional competitiveness, and a reasonable and coordinated innovation network structure can promote high-quality regional development. Utilizing the modified gravity model and social network analysis method, an innovation network composed of 27 cities in the Yangtze River Delta urban agglomeration from 2010 to 2021 was studied. The following conclusions were founded: (1) The innovation development level in the Yangtze River Delta urban agglomeration was constantly improving, and the innovation development level generally showed a spatial pattern of high in the southeast and low in the northwest. (2) The intensity and density of innovation network correlations in urban agglomerations were increasing, and the centrality of network nodes had an obvious hierarchical characteristic. The innovation network had a significant core–periphery spatial structure, with core cities that had higher centrality, such as Shanghai, Nanjing, and Hangzhou, playing the role of “intermediaries” and “bridges”, while cities with lower centrality, such as Anhui and cities in northern Jiangsu, generally played the role of “periphery actors” in the network. (3) The spatial correlation network of innovation of the Yangtze River Delta urban agglomeration could be divided into four blocks, namely, main benefit, broker, two-way spillover, and net spillover, and the spillover effect among them had obvious gradient characteristics of hierarchy.

YOLO-ViT-Based Method for Unmanned Aerial Vehicle Infrared Vehicle Target Detection

The detection of infrared vehicle targets by UAVs poses significant challenges in the presence of complex ground backgrounds, high target density, and a large proportion of small targets, which result in high false alarm rates. To alleviate these deficiencies, a novel YOLOv7-based, multi-scale target detection method for infrared vehicle targets is proposed, which is termed YOLO-ViT. Firstly, within the YOLOV7-based framework, the lightweight MobileViT network is incorporated as the feature extraction backbone network to fully extract the local and global features of the object and reduce the complexity of the model. Secondly, an innovative C3-PANet neural network structure is delicately designed, which adopts the CARAFE upsampling method to utilize the semantic information in the feature map and improve the model’s recognition accuracy of the target region. In conjunction with the C3 structure, the receptive field will be increased to enhance the network’s accuracy in recognizing small targets and model generalization ability. Finally, the K-means++ clustering method is utilized to optimize the anchor box size, leading to the design of anchor boxes better suited for detecting small infrared targets from UAVs, thereby improving detection efficiency. The present article showcases experimental findings attained through the use of the HIT-UAV public dataset. The results demonstrate that the enhanced YOLO-ViT approach, in comparison to the original method, achieves a reduction in the number of parameters by 49.9% and floating-point operations by 67.9%. Furthermore, the mean average precision (mAP) exhibits an improvement of 0.9% over the existing algorithm, reaching a value of 94.5%, which validates the effectiveness of the method for UAV infrared vehicle target detection.

Innovation network structure, government R&D investment and regional innovation efficiency: Evidence from China.

Based on the panel data of 30 provinces in China from 2011 to 2019, this paper uses a two-stage DEA model to measure regional innovation efficiency, then non-parametric test is used to examine the impact of innovation network structure and government R&D investment on regional innovation efficiency. The results show that, at the provincial level, innovation efficiency of regional R&D is not necessarily in direct proportion to the innovation efficiency in the commercialization stage. Commercialization efficiency is not necessarily high in provinces with high technical R&D efficiency. At the national level, the innovation efficiency gap between our country's R&D and commercialization stage is small, indicating that the development of the national innovation efficiency is more and more balanced. Innovation network structure can promote the R&D efficiency, but has no significant effect on the commercialization efficiency. Government R&D investment helps to improve the R&D efficiency, but it is not conducive to the improvement of commercialization efficiency. The interaction between innovation network structure and government R&D investment will have compound effects on regional innovation efficiency; the region with underdeveloped innovation network structure can increase the government R&D investment to make it have a higher level of R&D. This paper provides insights into how to improve innovation efficiency in different social networks and policy environments.

U²-Net for 3D Electrical Impedance Tomography With Combined Electrodes

A flexible Electrical Impedance Tomography (EIT) sensor with combined electrodes is proposed to reconstruct 3D image. The 3D EIT sensor consists of 24 combined electrodes, each consisting of a center electrode and an around electrode. The accuracy of measurement data has been improved by using the combined electrodes. In this study, an innovative U2-Net neural network structure is implemented to solve the inverse problem of 3D EIT. Significant amounts of priori information obtained from simulation are stored in the U2-Net, which facilitates the improvement of image reconstruction quality. The proposed 3D EIT sensor is verified by simulation and experiment. In the simulations, U2-Net shows better performance than U-Net. Combined electrodes show better robustness than conventional structure which shows higher accuracy when the signal to noise ratio (SNR) is 20 dB or 30 dB. In the experiments, structural similarity index measure (SSIM) is proposed to evaluate the image reconstruction quality. Experimental results show that the proposed 3D EIT sensor using U2-Net improves accuracy of image reconstruction. In addition, the proposed 3D EIT sensor improves the robustness of measurement results, which is satisfactory for the requirements of 3D image reconstruction.

Digital Development Influencing Mechanism on Green Innovation Performance: A Perspective of Green Innovation Network

Digital development offers groundbreaking opportunities for environmental monitoring and protection, as well as for maintaining the overall health of the planet. Through judicious use, global sustainability can be promoted. This study constructs a green innovation network using the green innovation patent data of listed companies of new energy vehicles as empirical evidence. The influence of network structure on green innovation performance is analyzed, and network structure indicators with positive effects are identified. The Exponential Random Graph Model (ERGM) is used to analyze the evolutionary influence on the structure of green innovation network from network configuration as endogenous attributes and digital characteristics as exogenous attributes. We explore the mechanism of improving green innovation performance by digitalization through enhancing the development of green innovation network. The results show that the enhanced structure of innovation network has a significant positive effect on the improvement of green innovation performance. Both "star" and "closed triangle" network configurations can promote the establishment of green innovation cooperation. In terms of digital characteristics, digital technology and digital organization can positively promote the development of green innovation network, while digital environment does not show a significant facilitating effect. Because of the tremendous expansion of digital technology adoption and the necessary infrastructure, enterprises can bridge the technology gap by optimizing digital technology and digital organizations. Enterprises can improve green innovation performance by optimizing digital technology and digital organization to promote the establishment of green innovation cooperation. The government should avoid the ineffective empowerment of digital environment construction on green innovation performance.

Impact of innovation network on regional innovation performance: do network density, network openness and network strength have any influence?

Purpose With the development of regional economy, innovation network plays an increasingly prominent role in reducing regional innovation cost and enabling information, knowledge and capital flow and diffusion. Building an efficient innovation network has become a feasible way to improve regional innovation capabilities and performance. Thus, under the background of Chinese special triple helix structure, the specific mechanism of innovation network characteristics on regional innovation performance is still an issue that needs to be studied urgently. Design/methodology/approach This study conducts a multi-level regression analysis regional innovation panel data of China from the past four years to explore how the three dimensions of innovation network characteristics (i.e. network density, openness and strength) affect regional innovation performance. Findings The results show that there is an inverted-U relationship exists between network density and regional innovation performance. The network openness and regional innovation performance show a significantly positive relationship and a U-shaped relationship exists between network strength and regional innovation performance. Research limitations/implications First, this study examines the relationship between network density, openness, strength and innovation ability in the network variable structure. However, this study does not analyze how absorptive capacity impacts the network structure and innovation performance of regional innovators. Second, innovation network intensity largely varies according to different types of enterprises or industries. Therefore, future studies can attempt to analyze the relationship between innovation network and innovation performance on the basis of the industry or the enterprise itself. Fourth, this study does not consider the change in the influence of innovation network structure on innovation ability. Practical implications The results of this study provide insights for the formulation of the regional innovation policy. First, enterprises must maintain good contact with research institutes, universities and technology intermediaries and promote resource, information and money flow between networks through formal and informal interactions. Enterprises can currently only interact with a limited number of innovative subjects due to their limited resources. Therefore, we should continue implementing the open policy of foreign capital utilization and absorb the technology, knowledge, management, ideas, talent and other resource advantages in the world. The investment environment in the central and western regions should be improved as soon as possible to guide foreign direct investments to the middle and the west part of China, thereby promoting the coordinated development of regional innovation ability and economy. Social implications In 2015, the Chinese Government proposed “public entrepreneurship and innovation,” including technological, institutional, management and mode innovations. This study suggests that innovation subjects in the region should establish an interactive and dynamic innovation network among innovators. Originality/value The innovation of this study lies in the analysis of the characteristics of innovation network that significantly affect regional innovation performance, the exploration of different stages of innovation network construction path and provide theoretical guidance for the construction of innovation network.

Evolution of innovation networks in industrial clusters and multidimensional proximity: A case of Chinese cultural clusters

The innovation and optimization of cultural industry clusters is important to the promotion of economic structural transformation and high-quality development. Taking 31 provinces (municipalities directly under the Central Government and autonomous regions in China) as the research object, the cross-sectional data included the number of joint patent applications for culture and related industries among provinces (regions) in 2014, 2016, 2018 and 2020. Then, the cluster innovation network was constructed using social network analysis (SNA) based on scientific identification of cultural industry clusters. This new cluster innovation network structure and its spatial evolution characteristics are accessible with the help of UCINet and ArcGIS tools. Finally, the multiple regression-quadratic assignment procedure (MR-QAP) method is used to determine the influence of geographical proximity, technological proximity, and social proximity on the evolution pattern of the cultural industry cluster innovation network. The results of the study are important for improving the efficiency of national cultural industry innovation and promoting regional synergistic development.

An Empirical Analysis on the Impact of Innovation Network Structure on Crossover Innovation Performance of Emerging Technologies.

The crossover innovation springing up in emerging technologies has drawn wide attention from scholars. Innovation network, as an effective way for major innovation-driven entities towards less relevant risks and higher efficiency, can significantly affect the crossover innovation performance. This paper analyzes the evolution law of the innovation network of autonomous driving technology based on the Social Network Analysis (SNA) and by using the data on joint applications for invention patents of such technology during 2006–2020. Furthermore, the structural eigenvalues of the network evolution are calculated for the regression analysis of the relationship between network structure and crossover innovation performance. The empirical results show that network centrality, structural hole, and relationship intensity have a positive effect on crossover innovation performance of emerging technologies, while network clustering has a negative effect. Emerging technology enterprises should constantly improve their technological innovation ability, improve their status and influence in the innovation network, establish cooperation with appropriate innovation partners, further expand their own technical knowledge fields, and obtain innovation resources by optimizing the network structure so as to enhance the crossover innovation performance.

Formation and Evolution of Ideal Interfirm Collaborative Innovation Networks Based on Decision-Making Rules for Partner Selection

On the basis of an external and static perspective on the topological structure of collaborative innovation networks, it is extremely difficult to answer the two most important concerns, namely, which structure is ideal and how to develop it in practice. By contrast, this study transfers to internal and dynamic perspectives, and then proposes that the essence of developing the ideal network lies in choosing the best partners. Therefore, we firstly propose the basic decision-making rules for selecting partners. In order of priority: knowledge distance, knowledge complementarity and barter exchange. Secondly, a model is constructed to describe this process of selecting partners and exchanging knowledge. Thirdly, the simulation results show that a small-world network is ideal in the initial stage of collaborative innovation. However, a random network is ideal in the mature periods. This result shows that the ideal network structure is not fixed, but affected by the life cycle of collaborative innovation alliance. Furthermore, this supports the notion that a small world is spontaneously generated in the real world, and also confirms that the formation of a small-world network will be driven intrinsically by a firm’s demand for external knowledge, and not necessarily by the external driving force of social capital. Finally, these findings solve the above two most important questions.

Research on the Governance Mechanism of Independent Innovation Network in the Core Area of Silk Road Economic Belt

The increasing competition in global scientific and technological innovation and the continuous promotion of the “Belt and Road initiatives” have created new opportunities for the improvement in innovation capabilities in the core area of the Silk Road Economic Belt but have also brought new challenges. This paper first starts from the market and government perspective and sorts out the formation mechanism of an innovation network. Subsequently, based on the development practice of the innovation network in the core area of the Silk Road Economic Belt, we applied the grounded theory to reveal the development constraints of the innovation network in the core area and explore the corresponding network governance mechanism. Subsequently, we applied a survey study to test our conceptual model. We empirically found the innovation network structure governance and innovation relation governance impacting mechanism in a market-led and government guidance context. The study found that: first, the market-led environment and government guidance jointly impact the formation of an innovation network. Second, the network structure governance and network relation governance of the innovation network can address the development constraints faced by the current innovation network to a certain extent, thereby improving the firm’s innovation performance. Finally, it is found that both the market-led environment and government guidance play a positive role in regulating the impact of network structure governance and network relation governance on the firm’s innovation performance. The above conclusions expand the application scenarios of innovation network governance theory in underdeveloped areas and, at the same time, provide new theoretical knowledge increments for the formation mechanism of innovation networks and also provide corresponding theoretical guidance for the improvement of the firm’s innovation performance in the core area.

The effects of proximities on the evolving structure of intercity innovation networks in the Guangdong–Hong Kong–Macao Greater Bay Area: comparison between scientific and technology knowledge

ABSTRACT The various knowledge flows shape and change the regional innovation patterns, which are also influenced by regional conditions. What are the similarities, differences and connections between science and technology linkage, as two different types of knowledge network, deserve to be explored in depth. Drawing on scientific paper co-publications and patent transfer data, we constructed two different types of intercity innovation networks during 2006–2016 in the Guangdong–Hong Kong–Macao Greater Bay Area (GBA), a city region special for its ‘one country, two systems’ structure. After that, we explored the evolutionary characteristics of the networks and further examined the different impacts of multidimensional proximity on scientific collaboration and technology transfer. Our results show that technology transfer is more sensitive to spatial factors, institutional barriers caused by ‘one country, two systems’ is a bigger obstacle to technology transfer between cities, cultural proximity and cognitive proximity have a more significant impact on paper cooperation network. Moreover, geographical proximity can indirectly affect knowledge spillover by acting on the proximity of other dimensions. As for scientific collaboration, social, cognitive and institutional proximities can compensate for the lack of geographical proximity, and cultural proximity frequently goes along with geographical proximity; as for technology transfer, geographical proximity has neither substitutional or complementarity relations with cultural and cognitive proximities, the interrelatedness between geographical and institutional and social proximities are complementarity which is opposed to paper co-publications. This study explores the differences in spillover mechanisms of different knowledge types and contributes to enriching the empirical framework of multidimensional proximities and innovation network researches.

Spatial Pattern and Evolution of Global Innovation Network from 2000 to 2019: Global Patent Dataset Perspective

In the era of the knowledge economy, the improvement of national innovation systems is playing a significant role in the global entrepreneurship ecosystem. Entrepreneurs are accelerating international intellectual property applications to be competitive. What remains to be explored is the evolution of international intellectual property network in the globe. With the application of social network analysis and intellectual property application database, the global innovation network structure from 2000 to 2019 is explored. Results showed that (1) in the period 2000–2019, the global innovation network has been expanding rapidly from a sparse network to a dense and complex one. (2) Patent application is unevenly distributed in the globe. Countries such as the US, China, and Canada have been the top countries flowing in, while Japan, Korea, EU, and Switzerland have been the main countries flowing out. (3) Global innovation network shows an obvious “core-periphery” pattern. The distribution pattern presents a quadrilateral structure with the four core regions of “US, Japan, EU, and China” as the apex. This analysis contributes to the visualization of the global layout of intellectual property and the evolution trend by analyzing intellectual property application networks. This can provide important experience reference for enterprises to study the global entrepreneurship ecosystem.

KIEGLFN: A unified acne grading framework on face images

Background and Objective: Grading the severity level is an extremely important procedure for correct diagnoses and personalized treatment schemes for acne. However, the acne grading criteria are not unified in the medical field. This work aims to develop an acne diagnosis system that can be generalized to various criteria. Methods: A unified acne grading framework that can be generalized to apply referring to different grading criteria is developed. It imitates the global estimation of the dermatologist diagnosis in two steps. First, an adaptive image preprocessing method effectively filters meaningless information and enhances key information. Next, an innovative network structure fuses global deep features with local features to simulate the dermatologists’ comparison of local skin and global observation. In addition, a transfer fine-tuning strategy is proposed to transfer prior knowledge on one criterion to another criterion, which effectively improves the framework performance in case of insufficient data. Results: The Preprocessing method effectively filters meaningless areas and improves the performance of downstream models.The framework reaches accuracies of 84.52% and 59.35% on two datasets separately. Conclusions: The application of the framework on acne grading exceeds the state-of-the-art method by 1.71%, reaches the diagnostic level of a professional dermatologist and the transfer fine-tuning strategy improves the accuracy of 6.5% on the small data.