Data-driven Approach Research Articles

The deep continual learning framework for prediction of blast-induced overbreak in tunnel construction

Blast-induced overbreak, characterized by the excessive removal of rock mass beyond the planned tunnel profile, poses significant safety risks, increases costs, and causes project delays during tunneling. Traditional machine/deep learning models have been developed to predict overbreak. However, these models are often inadequate because they are static and lack the flexibility to adapt to new, real-world data continuously. This study addresses this limitation by introducing a novel data-driven approach based on deep continual learning. The primary objective is to develop an adaptable predictive model with the ability of continual learning, which is particularly advantageous in dynamic environments like tunnel blasting. To achieve this, a self-attention multi-layer perceptron (MLP) model for overbreak prediction, integrated with two continual learning strategies (elastic weight consolidation (EWC) and memory replay (MR)), is developed. This step enables the overbreak prediction model to possess the ability to continuously learn real-world scenarios and adapt to the dynamic environment of tunnel blasting. The findings show that the continuous MLP model, empowered by EWC and MR, demonstrates superior adaptability and accuracy in predicting overbreak. Compared with the standard MLP model, which achieves a predictive accuracy of 0.831, the continuous MLP model achieves a predictive accuracy of 0.845 on unseen data. The integration of EWC and MR strategies proves to be a pivotal factor in developing deep learning models for the dynamic task of predicting overbreak. The continual learning strategies ensure that the models remain adaptable and accurate over time, which is essential for practical applications in dynamic environments of tunnel blasting operations.

Evaluation of Machine Learning Models for Predicting the Hot Deformation Flow Stress of Sintered Al–Zn–Mg Alloy

Abstract In predicting flow stress, machine learning (ML) offers significant advantages by leveraging data-driven approaches, enhancing material design, and accurately forecasting material performance. Thus, the present study employs various supervised ML models, including linear regression (Lasso and Ridge), support vector regression (SVR), ensemble methods (random forest (RF), gradient boosting (GB), extreme gradient boosting (XGB)), and neural networks (artificial neural network (ANN), multilayer perceptron (MLP)), to predict flow stress in the hot deformation of an Al–Zn–Mg alloy. The ML methodology involves sequential steps from data extraction to cross-validation and hyperparameter tuning, which is conducted using the hyperopt library. Model performance is assessed using average absolute relative error (AARE), root-mean-squared error (RMSE), and mean squared error (MSE). The results show that ensemble methods (RF, GB, XGB) and neural networks outperform traditional regression methods, demonstrating superior predictive accuracy. Visualization using half-violin plots reveals the models' error ranges, with XGB consistently exhibiting the best performance. SVR, RF, GB, XGB, ANN, and MLP showed better performance than the Arrhenius model in the context of AARE and MSE metrics. Interestingly, SVR had a somewhat higher AARE of 1.89% and an MSE of 0.251 MPa2, while XGB had the lowest AARE of 0.2% and the lowest MSE of 0.011 MPa2. When ML models were evaluated using the skill score in relation to the Arrhenius model, XGB scored higher than the support vector machine (SVM) at 0.714, with a score of 0.986. In contrast, Lasso and Ridge exhibited negative scores of −0.847 and −0.456, respectively.

A Fuzzy Gold Mineral Prediction Model Integrating with Knowledge-Driven and Data-Driven: A Case Study of the Hatu Region in Xinjiang, China

As mineral resources become increasingly scarce, the search for potential ore deposits presents a significant challenge in geological exploration. Subjective factors often constrain traditional knowledge-driven approaches, while purely data-driven methods may overlook the geological significance of data relationships, potentially compromising the accuracy and reliability of predictions. To address these issues, this study proposes a novel mineral prediction model that integrates fuzzy comprehensive evaluation with both knowledge-driven and data-driven approaches (FCEKDD), resulting in an optimized comprehensive mineral prediction model based on multi-source geoscience data. This model establishes comprehensive exploration indicators based on expert knowledge and quantitatively assesses these indicators through fuzzy evaluation methods to evaluate mineralization potential, thereby delineating exploration targets within the study area. Ultimately, a three-layer predictive framework is constructed using the C-A fractal method. Experimental results indicate that 57.1% of known mineral points are within the primary prediction zone, thus demonstrating the model’s high predictive accuracy. Comparisons with a random forest model reveal that the FCEKDD model has advantages in addressing geoscience data uncertainty and interpreting geological phenomena. This research validates the reliability and applicability of the proposed model in mineral exploration mapping, providing an effective solution for future mineral resource exploration.

Bifurcation analysis in dynamical systems through integration of machine learning and dynamical systems theory

Abstract Characterizing the nonlinear behavior of dynamical systems near the stability boundary is a critical step toward understanding, designing, and controlling systems prone to stability concerns. Traditional methods for bifurcation analysis in both experimental systems and large-dimensional models are often hindered either by the absence of an accurate model or by the analytical complexity involved. This paper presents a novel approach that combines the theoretical frameworks of nonlinear reduced-order modeling and stability analysis in dynamical systems with advanced machine learning techniques to perform bifurcation analysis in dynamical systems. By focusing on a low-dimensional nonlinear invariant manifold, this work proposes a data-driven methodology that simplifies the process of bifurcation analysis in dynamical systems. The core of our approach lies in utilizing carefully designed neural networks to identify nonlinear transformations that map observation space into reduced manifold coordinates in its normal form where bifurcation analysis can be performed. The unique integration of analytical and data-driven approaches in the proposed method enables the learning of these transformations and the performance of bifurcation analysis with a limited number of trajectories. Therefore, this approach improves bifurcation analysis in model-less experimental systems and cost-sensitive high-fidelity simulations. The effectiveness of this approach is demonstrated across several examples.

Time Series Analysis: A Model Solution for Addressing Nigeria's Economic Challenges

This study undertakes a comprehensive time series analysis of Nigeria's economic data from 2010 to 2023, focusing on key indicators such as GDP, inflation rate, unemployment rate, exchange rate, foreign direct investment (FDI), government revenue, and expenditure. Utilizing the ARIMA (Autoregressive Integrated Moving Average) model, we aim to forecast future economic trends and provide actionable insights for policy-making. Our analysis reveals significant volatility in Nigeria's economic indicators over the studied period. The GDP showed fluctuations, peaking in 2014 before experiencing a decline due to the 2016 recession, followed by gradual recovery. The inflation rate remained high, peaking at 19.5% in 2023, reflecting persistent inflationary pressures. Unemployment showed a troubling upward trend, reaching a height of 35.0% in 2021, indicating structural economic issues. The exchange rate exhibited substantial depreciation, while FDI inflows showed a declining trend, highlighting challenges in attracting foreign investment. Correlation analysis revealed critical relationships between these indicators: higher inflation rates and unemployment are associated with lower GDP growth, while higher exchange rates (weaker currency) associate with higher inflation. The ARIMA model forecast for GDP suggests continued recovery, but with significant uncertainties. The findings underscore the need for diversified economic strategies, effective inflation control measures, targeted job creation policies, and improved conditions for FDI. These insights are crucial for policymakers aiming to stabilize and grow Nigeria's economy. This study highlights the importance of data-driven approaches in addressing economic challenges and promoting sustainable growth in Nigeria.

Born Connected: Do Infants Already Have Adult-Like Multi-Scale Connectivity Networks?

The human brain undergoes remarkable development with the first six postnatal months witnessing the most dramatic structural and functional changes, making this period critical for in-depth research. rsfMRI studies have identified intrinsic connectivity networks (ICNs), including the default mode network, in infants. Although early formation of these networks has been suggested, the specific identification and number of ICNs reported in infants vary widely, leading to inconclusive findings. In adults, ICNs have provided valuable insights into brain function, spanning various mental states and disorders. A recent study analyzed data from over 100,000 subjects and generated a template of 105 multi-scale ICNs enhancing replicability and generalizability across studies. Yet, the presence of these ICNs in infants has not been investigated. This study addresses this significant gap by evaluating the presence of these multi-scale ICNs in infants, offering critical insight into the early stages of brain development and establishing a baseline for longitudinal studies. To accomplish this goal, we employ two sets of analyses. First, we employ a fully data-driven approach to investigate the presence of these ICNs from infant data itself. Towards this aim, we also introduce burst independent component analysis (bICA), which provides reliable and unbiased network identification. The results reveal the presence of these multi-scale ICNs in infants, showing a high correlation with the template (rho > 0.5), highlighting the potential for longitudinal continuity in such studies. We next demonstrate that reference-informed ICA-based techniques can reliably estimate these ICNs in infants, highlighting the feasibility of leveraging the NeuroMark framework for robust brain network extraction. This approach not only enhances cross-study comparisons across lifespans but also facilitates the study of brain changes across different age ranges. In summary, our study highlights the novel discovery that the infant brain already possesses ICNs that are widely observed in older cohorts.

Avaliações de Usuários do Aplicativo "Gov.br" como Base para Análise de Usabilidade

This article presents an analysis of the usability of the "Gov.br" app, based on user comments left on the Google Play Store, aiming to identify areas for improvement in the user experience. The central question of the study is how user feedback can contribute to enhancing the application. This is a descriptive study grounded in usability heuristics found in the scientific literature. The results show that user comments highlight significant issues, such as functionality errors, display difficulties, and overly complex tasks. The suggestions emphasize the need to simplify processes, increase accessibility, and personalize features. Comments with intermediate ratings (between 2 and 4 stars) proved particularly useful for identifying specific issues and proposing solutions. The analysis of these opinions emerges as an effective approach for evaluating and improving the usability of government mobile applications. The study underscores the relevance of a data-driven approach to guide the development of user-centered interfaces.

A data-driven hybrid approach towards developing a circular economy diffusion model for the building construction industry

Although there is a growing body of knowledge about circular economy (CE) practices in the building construction industry (BCI), the prediction of CE diffusion within the BCI remains insufficiently explored. This paper aims to explore a hybrid approach towards predicting the diffusion of CE practices in the BCI of a developing economy. This study utilized the technology-organization-environment (TOE) framework to identify the essential factors influencing CE diffusion in the BCI. Survey data collected from 303 experts were analyzed using partial least squares structural equation modelling (PLS-SEM) to test the hypothesis regarding these influencing factors. Subsequently, machine learning (ML) algorithms were employed to develop a predictive model for CE diffusion in the BCI. SHapley Additive exPlanation (SHAP) was then applied to interpret the contributions of each essential factor to the predictive model. The PLS-SEM results advocated that four major factors, namely technological compatibility, relative technological advantages, top management support, and organizational readiness, significantly and positively influence CE diffusion in the BCI. Furthermore, random forest algorithm was identified as the optimal ML model for predicting CE diffusion, achieving an accuracy of 80.33% and ROC AUC (area under the Curve receiver operating characteristics) of 80.27%. According to the SHAP results, the three most essential features contributing to the random forest model prediction are organizational readiness, top management support, and the relative technological advantages of CE adoption. This study advances the existing literature on CE diffusion by offering a comprehensive, data-driven approach that stakeholders can leverage to forecast trends and patterns in CE practices. It also equips decision makers with strategic insights and pragmatic plans to foster CE diffusion in the BCI, particularly within the context of developing countries.

Intelligent Fault Diagnosis Across Varying Working Conditions Using Triplex Transfer LSTM for Enhanced Generalization

Fault diagnosis plays a pivotal role in ensuring the reliability and efficiency of industrial machinery. While various machine/deep learning algorithms have been employed extensively for diagnosing faults in bearings and gears, the scarcity of data and the limited availability of labels have become a major bottleneck in developing data-driven diagnosis approaches, restricting the accuracy of deep networks. To overcome the limitations of insufficient labeled data and domain shift problems, an intelligent, data-driven approach based on the Triplex Transfer Long Short-Term Memory (TTLSTM) network is presented, which leverages transfer learning and fine-tuning strategies. Our proposed methodology uses empirical mode decomposition (EMD) to extract pertinent features from raw vibrational signals and utilizes Pearson correlation coefficients (PCC) for feature selection. L2 regularization transfer learning is utilized to mitigate the overfitting problem and to improve the model’s adaptability in diverse working conditions, especially in scenarios with limited labeled data. Compared with traditional transfer learning approaches, such as TCA, BDA, and JDA, which demonstrate accuracies in the range of 40–50%, our proposed model excels in identifying machinery faults with minimal labeled data by achieving 99.09% accuracy. Moreover, it performs significantly better than classical methods like SVM, RF, and CNN-based networks found in the literature, demonstrating the improved performance of our approach in fault diagnosis under varying working conditions and proving its applicability in real-world applications.

U + LSTM-F: A data-driven growth process model of rice seedlings

Accurately predicting the growth status of rice seedlings and understanding their growth rate and health status in a timely manner helps adjust the growth cycle and management measures. By predicting the growth status of the seedlings, the best time for transplanting can be selected, improving the survival rate and overall health of the seedlings, thereby enhancing yield and quality. Therefore, this study proposes a data-driven time-series model, the U + LSTM-F model, for predicting the growth status of Wuyou Rice 4 seedlings. First, the U-Net model is employed to segment sequentially collected images, extracting features such as leaf age and stem length of the rice seedlings. Subsequently, the collected ambient temperature and humidity data are aligned with the leaf age and stem length data. Finally, the LSTM model is used for time-series analysis, enabling the model to learn the temporal relationship between environmental and growth data and predict the growth trend of the rice seedlings. Additionally, an attention mechanism is introduced to enhance model performance, and the model's effectiveness is evaluated using multiple quantitative metrics. The proposed model achieves an RMSE of 0.032 and MAPE of 0.895 % for leaf age prediction, and an RMSE of 0.067 and MAPE of 0.814 % for stem length prediction. The experimental results show that this data-driven approach, which combines growth data with environmental data, exhibits high accuracy in predicting the leaf age and stem length of rice seedlings. This provides a more accurate tool for predicting the growth of rice seedlings, offering valuable insights for rice seedling cultivation research.

Advancements of Exploiting Convolutional Neural Networks for Solving Differential Equations

Solving Partial Differential Equations (PDEs) is essential across various fields, including physics, mathematics, and engineering. This study explores innovative methods for solving PDEs using Convolutional Neural Networks (CNNs). Traditionally, solving PDEs through numerical methods like finite difference or spectral approaches is computationally intensive, particularly when addressing high-dimensional problems and complex boundary conditions. CNNs, with their ability to handle spatial data efficiently, offer a promising alternative. This research evaluates different neural network architectures, including data-driven, physics-driven, and hybrid approaches, and their role in enhancing the accuracy and scalability of PDE solutions. Data-driven methods rely on extensive datasets for training neural networks, while Physics-Informed Neural Networks (PINNs) integrate the underlying physical laws into the solution process. Hybrid approaches combine both strategies, leveraging data and physical principles. The paper emphasizes CNNs potential to transform traditional PDE-solving methods, making them more efficient and adaptable to complex, real-world applications. Future research directions include optimizing these methods for greater scalability and interdisciplinary applications.

A Receiver Position Estimation Method Based on LSTM for Multi-Transmitter Single-Receiver Wireless Power Transfer Systems

The multi-transmitter single-receiver wireless power transfer (MTSR-WPT) system has good tolerance for coil misalignment because the magnetic fields generated by multiple transmitters can be shaped to adapt to position changes in the receiver coil. In order to achieve magnetic field shaping of the MTSR-WPT system and increase power transfer efficiency, accurately estimating the position of the receiver coil is a key issue that needs to be addressed. In this article, a receiver position estimation method based on long short-term memory (LSTM) is proposed, which utilizes a data-driven approach to establish a neural network model. By learning the relationship between the measured time-series voltage data of the transmitter coils and the position of the receiver coil, the proposed model can achieve accurate position estimation of the receiver. Compared with previous works, the proposed method does not require communication between the transmitter and receiver, which is conducive to simplifying the system structure and reducing costs. In addition, the proposed LSTM-based method requires less derivation of complex formulas and the internal mechanism analysis of the system. Finally, a MTSR-WPT prototype is built to verify the proposed method. The experimental results show that the proposed LSTM-based method can achieve high-accuracy position estimation of the receiver. When the receiver moves within a range of 160 mm × 160 mm, the average error between the estimated receiver coil position using the proposed method and the actual receiver coil position is less than 2.40 mm.

A hybrid prognostic framework: Stochastic degradation process with adaptive trajectory learning to transfer historical health knowledge

Remaining useful life (RUL) prediction is crucial to supporting intelligent maintenance and health management of safety–critical products. Although advanced data-driven approaches such as deep neural networks are effective in processing high-dimensional non-linear health features, their application to field RUL prediction confronts with two challenges: (a) adaptivity of the lifetime parameter learning process is often restricted, and (b) prediction of multi-source uncertainties is almost analytically intractable. This paper addresses such challenges by devising a tractable, global adaptive model-data-interaction prognostic framework, where a non-linear stochastic degradation model governed by self-adaptive trajectory pattern is constructed to transfer historical health knowledge. In particular, a joint parameter learning framework is established under the structure of a multi-branch Bayesian network, such that to simultaneously learn: (a) degradation model parameters, and (b) network hyper-parameters. Additionally, the key control parameters of the degradation process are updated adaptively leveraging multi-dimensional sequential Bayesian learning. An efficient interpolation algorithm is further proposed to alleviate computation burden of RUL distributions. Case studies conducted on both turbofan engines degradation data and field train bearing vibration data demonstrate the superior model performance compared to existing methodologies.

Generalized representative structures for atomistic systems.

A new method is presented to generate atomic structures that reproduce the essential characteristics of arbitrary material systems, phases, or ensembles. Previous methods allow one to reproduce the essential characteristics (e.g., the chemical disorder) of a large random alloy within a small crystal structure. The ability to generate small representations of random alloys, along with the restriction to crystal systems, results from using the fixed-lattice cluster correlations to describe structural characteristics. A more general description of the structural characteristics of atomic systems is obtained using complete sets of atomic environment descriptors. These are used within for generating representative atomic structures without restriction to fixed lattices. A general data-driven approach is provided here utilizing the atomic cluster expansion (ACE) basis. TheN-body ACE descriptors are a complete set of atomic environment descriptors that span both chemical and spatial degrees of freedom and are used within for describing atomic structures. The generalized representative structure (GRS) method presented within generates small atomic structures that reproduce ACE descriptor distributions corresponding to arbitrary structural and chemical complexity. It is shown that systematically improvable representations of crystalline systems on fixed parent lattices, amorphous materials, liquids, and ensembles of atomic structures may be produced efficiently through optimization algorithms. With the GRS method, we highlight reduced representations of atomistic machine-learning training datasets that contain similar amounts of information and small 40-72 atom representations of liquid phases. The ability to use GRS methodology as a driver for informed novel structure generation is also demonstrated. The advantages over other data-driven methods and state-of-the-art methods restricted to high-symmetry systems are highlighted.

Research on the Risk Control Model Development Based on Scorecard

Abstract. Effective risk management is vital for ensuring the stability and profitability of financial institutions. This study focuses on enhancing credit risk assessment by developing a scoring model that quantifies customer risk factors using logistic regression. Key features such as Weight of Evidence (WOE) and Information Value (IV) were employed to transform and select variables. The datasets used in this research, cs-training.csv and cs-test.csv, were preprocessed, including handling missing data and binning continuous variables to improve model interpretability and performance. The logistic regression model yielded an AUC value of 0.78, demonstrating strong predictive capabilities, though optimization is necessary to improve the F1 score. The final scorecard generated from the model provides actionable insights for financial institutions, enabling more accurate risk predictions and decision-making. This tool is particularly useful for loan approvals and credit issuance, offering a data-driven approach to managing credit risk in modern financial environments.

Credit Evaluation System Based on FICO

Abstract. As credit transactions become more prevalent, financial institutions require effective methods to assess credit risk and reduce the likelihood of borrower default. In the U.S., the Fair Isaac Credit Organization (FICO) score is widely used by banks and insurers to evaluate personal creditworthiness. This paper aims to develop an automated credit scoring tool based on the FICO system to help financial institutions improve risk assessment. The paper leverages the random forest algorithm for data preprocessing and feature engineering to extract key variables, such as the borrower's financial status and credit history. To ensure data stability and interpretability, Information Value and Weight of Evidence techniques are applied to process these variables. Additionally, the Sigmoid function is used to map the model output to a range between 0 and 1, making it suitable for generating credit scores. This random forest algorithm helps handle non-linear relationships and missing data, while cross-validation enhances the models generalization ability. After training, the paper achieved an automated credit scoring system, closely aligned with the FICO scoring system. The models Area Under the Curve (AUC) value reached 0.84, indicating strong predictive accuracy and reliability. This tool enables financial institutions to more accurately assess credit risk, offering a robust, data-driven approach to improve decision-making and risk management.

Optimising Physics-Informed Neural Network Solvers for Turbulence Modelling: A Study on Solver Constraints Against a Data-Driven Approach

Physics-informed neural networks (PINNs) have emerged as a promising approach for simulating nonlinear physical systems, particularly in the field of fluid dynamics and turbulence modelling. Traditional turbulence models often rely on simplifying assumptions or closed numerical models, which simplify the flow, leading to inaccurate flow predictions or long solve times. This study examines solver constraints in a PINNs solver, aiming to generate an understanding of an optimal PINNs solver with reduced constraints compared with the numerically closed models used in traditional computational fluid dynamics (CFD). PINNs were implemented in a periodic hill flow case and compared with a simple data-driven approach to neural network modelling to show the limitations of a data-driven model on a small dataset (as is common in engineering design). A standard full equation PINNs model with predicted first-order stress terms was compared against reduced-boundary models and reduced-order models, with different levels of assumptions made about the flow to monitor the effect on the flow field predictions. The results in all cases showed good agreement against direct numerical simulation (DNS) data, with only boundary conditions provided for training as in numerical modelling. The efficacy of reduced-order models was shown using a continuity only model to accurately predict the flow fields within 0.147 and 2.6 percentage errors for streamwise and transverse velocities, respectively, and a modified mixing length model was used to show the effect of poor assumptions on the model, including poor convergence at the flow boundaries, despite a reduced solve time compared with a numerically closed equation set. The results agree with contemporary literature, indicating that physics-informed neural networks are a significant improvement in solve time compared with a data-driven approach, with a novel proposition of numerically derived unclosed equation sets being a good representation of a turbulent system. In conclusion, it is shown that numerically unclosed systems can be efficiently solved using reduced-order equation sets, potentially leading to a reduced compute requirement compared with traditional solver methods.

DEVELOPMENT OF A CREDIT SCORING MODEL USING MACHINE LEARNING FOR COMMERCIAL BANKS IN VIETNAM

In an increasingly competitive financial market, developing an accurate and efficient credit scoring model is crucial for banks to make informed credit decisions, and to help customers obtain credit products that match their financial capabilities. This study aims to develop a credit scoring model for individual customers of Vietnamese banks using machine learning techniques. This article will use a set of financial and non-financial indicators as inputs for various machine learning models such as Logistic Regression, K-nearest Neighbor, Decision Tree, Random Forest, LightGBM, and Support Vector Method. The obtained results from these models will be carefully compared and evaluated to ultimately select the best credit scoring model for the bank. This study leverages Google Colab, a cloud-based platform, for comprehensive data analysis and model development, ensuring a robust and data-driven approach. This study hopes to provide a useful solution for Vietnamese banks in managing credit risk and improving their business performance.

Machine learning prediction of tipping in complex dynamical systems

Anticipating a tipping point, a transition from one stable steady state to another, is a problem of broad relevance due to the ubiquity of the phenomenon in diverse fields. The steady-state nature of the dynamics about a tipping point makes its prediction significantly more challenging than predicting other types of critical transitions from oscillatory or chaotic dynamics. Exploiting the benefits of noise, we develop a general data-driven and machine-learning approach to predicting potential future tipping in nonautonomous dynamical systems and validate the framework using examples from different fields. As an application, we address the problem of predicting the potential collapse of the Atlantic meridional overturning circulation, possibly driven by climate-induced changes in the freshwater input to the North Atlantic. Our predictions based on synthetic and currently available empirical data place a potential collapse window spanning from 2040 to 2065, in consistency with the results in the current literature. Published by the American Physical Society 2024

Designing a regional clinical service for people with early-onset type 2 diabetes in England.

To design a regional clinical service for people with early-onset type 2 diabetes (EOT2D) in Leicester, Leicestershire and Rutland (England). A literature search was undertaken to identify important considerations. A working group of key stakeholders was formed to design a triage system and service pathway. Electronic medical records (EMRs) were searched (15th November 2023) to assess feasibility of the pathway and adapt accordingly. A literature search identified important considerations: High risk of complications; large proportion from minority ethnic and socioeconomically deprived backgrounds; significant psychological burden; stigma and other social challenges; and misclassification and miscoding. Novel clinical risk criteria were developed, implementable in EMRs, to match intervention-intensity to clinical need. Specialist clinics were planned, one for people at the highest-clinical risk, another for women with adverse perinatal risk factors. A healthcare professional training package was developed to increase awareness of the unmet clinical needs of people with EOT2D and to upskill in provision of holistic care. Subsequent EMR searches supported the need for our service. Due to the large numbers with HbA1c ≥86mmol/mol (10.0%; n=299; 10.8% of total), these people were prioritised for clinic access. We opted for specialist nurse/educator support to practices with clustering of patients and to financially incentivise referrals from primary care into services. We showcase a service specifically for people with EOT2D based on the literature, a broad range of stakeholder involvement and utilising a locally-sourced data-driven approach. We further discuss areas for development and recommendations based on the challenges we encountered.