Integrate Machine Learning Methods Research Articles

FRI-021 An integrative machine learning method enhances postoperative liver failure risk assessment with preoperative indicators

FRI-021 An integrative machine learning method enhances postoperative liver failure risk assessment with preoperative indicators

Insights into the Tennis Court through Machine Learning: Analysis and Evaluation of the Wimbledon Men's Singles Final

In today's data-driven world of sports science, in-depth analysis of tennis match data is especially critical for improving athlete performance, refining training strategies, and enhancing the viewing experience. This study starts from all aspects of data and key features of the players during the game, using Stacking integrated machine learning methods, a kind of integrated algorithms through XGBoost, GBDT, CatBoost, and ExtraTrees to build prediction models, aiming to always keep an eye on the dynamics of the game, using the data to reveal the changes in the player's strength and to quantify the impact of the momentum on the player as well as the game's trend. The experimental results show that we have identified factors that influence changes in players' momentum, which can provide real-time feedback to coaches and players so that they can make data-based decisions during matches, opening up new avenues for improving tennis players' scoring ability and maximizing help for players to improve their winning percentage in matches.

Analysing lexical semantic changes in chinese educational texts by integrating machine learning methods

Analyzing Lexical Semantic Changes (LSC) in Educational Texts (ET) refers to examining how the meanings of words, terms, or phrases used in ET have evolved. It involves learning shifts in the semantic content, connotations, and language associations within educational resources such as textbooks, research articles, and instructional content. The analysis can reveal how educational models, pedagogical methods, and terminology have transformed in response to technological innovations, societal changes, and pedagogical developments. This analysis provides visions into the dynamic nature of educational discourse, helping researchers, educators, and policymakers understand how language has adapted to reflect changes in educational paradigms and the broader educational context. This research investigates the semantic analysis and classification performance within ET, employing the innovative Decision Tree + Feed Forward Neural Networks (DT + FFNNs) framework. This research shows the dynamic semantic relationships inherent in educational terminology by diverse semantic similarity measures and contextualized embeddings. It looks at how educational language changes to reflect changes in society, technology, and pedagogy. The study uses a DT + FFNN framework for semantic analysis and classification. The study uses several embeddings and semantic similarity metrics, and Spearman’s Correlation Coefficient (SCC) is employed to evaluate their effectiveness. This study highlights the DT + FFNN framework’s capacity to capture complex semantics in an educational setting and offers insights into the adaptive nature of educational discourse. SCC serves as a guiding metric, offering insights into the efficiency of several embeddings and measures. The findings show the pivotal role of fine-tuning in significantly enhancing the accuracy of DT + FFNNs across measures, revealing its remarkable potential in capturing semantics within an educational context.

Effect of structural parameters on compression performance of autoclaved aerated concrete: Simulation and machine learning

Autoclaved aerated concrete (AAC) has broad applications in civil engineering due to its lightweight, thermal and sound insulation. However, the relation between the complex random porous structures and the compression performances of AAC is complex, limiting the optimization and application of AAC. In this study, finite element simulation was implemented and validated based on the experimental results. The simulation results showed that the compressive strength of AAC increased by 31.7, 45.8, and 134% with the porosity decreasing from 70% to 40%, the average pore diameter decreasing from 1.5 to 0.5 mm, and the pore connectivity decreasing from 50% to 0, respectively. Then, based on the numerical dataset, integrated machine learning methods were implemented to rapidly predict the compressive properties of AAC and analyze the main factors affecting the compressive properties. The ML results showed that the CatBoost model had the best predictive performance based on the small dataset of simulation results, with an average relative error of 18% and 7% for compressive strength and modulus of AAC, respectively. The component modulus was the most important feature for predicting the compressive strength and modulus of AAC.

Optimization of decision support systems in agriculture using AI: an integrated approach

Contemporary agriculture faces critical challenges in terms of sustainability and efficiency, in the context of constantpopulation growth and the need for environmentally responsible agricultural practices. This study addresses theapplication of Artificial Intelligence (AI) in the optimization of Decision Support Systems (SSD) for agriculture, with theaim of improving crop management and resource utilization. By integrating machine learning methods andimplementing predictive models, crop productivity was analyzed under various environmental conditions. The resultsdemonstrate improved accuracy in yield prediction, with a 15% increase in water and fertilizer use efficiency compared toconventional techniques. The research also reveals a significant increase in scientific collaboration and publication in thisfield since 2018, highlighting the importance of AI in the future of sustainable agriculture. The study's findings underlinethe potential of AI-based SSDs to transform agriculture, promoting more efficient and sustainable practices that canadapt to global challenges.

Short-term forecasting of streamflow by integrating machine learning methods combined with metaheuristic algorithms

This paper aims to introduce an artificial intelligence approach to enhance short-term daily streamflow forecasts through the integration of machine learning (ML) techniques with meta-heuristic (MH) algorithms. To maximize performance, a comprehensive analysis involving the autocorrelation and partial autocorrelation functions of the observed streamflow was conducted to derive ten distinct input combinations, from which the most suitable set was selected for investigation. Subsequently, six distinct ML methods, namely, decision trees, support vector regression, long short-term memory, gated recurrent unit, multi-layer perceptron, and radial basis function (as the optimal input combination), were leveraged to predict one-day-ahead streamflow. The most promising outcomes from these six explored methods were synergistically amalgamated to enhance predictive precision. This amalgamation was then optimized by utilizing parameter values hyper-tuned by four different MH algorithms. The algorithms included the shuffled frog leaping algorithm, particle swarm optimization, ant colony optimization (ACO), and gray wolf optimization (GWO). In culmination, a case study centered around the “Wick” river in the UK was conducted to rigorously evaluate the performance of the proposed method, utilizing a range of statistical evaluation indices. The results underscored the efficacy of the proposed combined method in substantially refining streamflow predictions, exhibiting an impressive level of accuracy. Additionally, the outcomes emphasized that ACO and GWO stand out as the optimal algorithms for calibrating the parameters of the amalgamated model. For instance, the R2 values attained by the combined models optimized through the ACO and GWO algorithms were 0.69422 and 0.69444, respectively. These values notably surpass those observed in the initial models, substantiating the superiority of the applied optimization approaches.

Integrated machine learning methods with oversampling technique for regional suitability prediction of waste-to-energy incineration projects

China's tiered strategy to enhance county-level waste incineration for energy aligns with the sustainable development goals (SDGs), emphasizing the need for comprehensive assessments of waste-to-energy (WtE) plant suitability. Traditional assessment methodologies face challenges, particularly in suggesting innovative site alternatives, adapting to new data sets, and their dependence on strict assumptions. This study introduced enhancements in three pivotal dimensions. Methodologically, it leverages data-driven machine learning (ML) approaches to capture the complex relationships essential for site selection, reducing dependency on strict assumptions. In terms of predictive performance, the integration of oversampling with stacked ensemble models enhances the diversity and generalizability of ML models. The area under curve (AUC) scores from four ML models, enhanced by the oversampled dataset, demonstrated significant improvements compared to the original dataset. The stacking model excelled, achieving a score of 92%. It also led in overall Precision and Recall, reaching 85.2% and 85.08% respectively. Nevertheless, a noticeable discrepancy existed in Precision and Recall for positive classes. The stacking model topped Precision scores at 83.1%, followed by eXtreme Gradient Boosting (XGBoost) (82.61%). In terms of Recall, XGBoost recorded the lowest at 85.07%, while the other three classifiers all marked 88.06%. From an industry applicability standpoint, the stacking model provides innovative location alternatives and demonstrates adaptability in Hunan province, offering a reusable tool for WtE location. In conclusion, this study not only enhances the methodological aspects of WtE site selection but also provides practical and adaptable solutions, contributing positively to sustainable waste management practices.

Plasma metabolomics and proteomics reveal novel molecular insights and biomarker panel for cholelithiasis

BackgroundCholelithiasis is a gastrointestinal disease that is associated with the highest socioeconomic cost. A diagnosis of cholelithiasis based on clinical features is significantly limited, and direct molecular insights into cholelithiasis and the relationship between cholelithiasis and clinical biochemical parameters are unclear. ObjectivesUncovering direct molecular insights into cholelithiasis and the relationship between cholelithiasis and clinical biochemical parameters. Identifying sensitive and specific biomarkers for this disease. MethodsParallel metabolomic and proteomic analyses of plasma from cholelithiasis patients (CPs) and healthy control individuals (HCs) without cholelithiasis were performed using ultrahigh-performance liquid chromatography-tandem mass spectrometry. A multimodule coexpression network analysis and integrated machine learning methods, including least absolute shrinkage and selection operator, random forest, and support vector machine, were used for bioinformatic analyses. An independent cohort and the cross-validation of the combination of two cohorts were used to evaluate the diagnostic performance of the panel. ResultsArachidonic acid metabolism was significantly different between the CP and HC groups. Glyceraldehyde-3-phosphate dehydrogenase, actin beta, phosphoglycerate mutase 1, Enolase 1, Myeloperoxidase, and actin alpha 1 were identified as potential proteins related to cholelithiasis. The correlation between the merged modules and clinical biochemical tests was calculated. A diagnostic panel composed of four candidate biomarkers, including 3-oxotetradecanoic acid, 12-hydroxydodecanoic acid, hemoglobin subunit delta (HBD), and fibrinogen beta chain (FGB), was proposed based on three modules that were significantly associated with cholelithiasis. The classification according to the diagnostic panel detected CPs with an area under the curve (AUC) of 0.955. External validation in an independent cohort resulted in similar accuracy (AUC=0.995). ConclusionsThis study provided some direct molecular insights into cholelithiasis by showing the differences in plasma metabolic and protein profiles between CPs and HCs and presented a potential biomarker panel with two metabolites (3-oxotetradecanoic acid, 12-hydroxydodecanoic acid) and two proteins (HBD, FGB) for predicting cholelithiasis. We also explored the potential correlation of clinical biochemical parameters with combined modules. These findings may provide some reference for the diagnosis of cholelithiasis in clinical practice.

Auxiliary guidance manufacture and revealing potential mechanism of perovskite solar cell using machine learning

To promote the development of global carbon neutrality, perovskite solar cells (PSCs) have become a research hotspot in related fields. How to obtain PSCs with expected performance and explore the potential factors affecting device performance are the research priorities in related fields. Although some classical computational methods can facilitate material development, they typically require complex mathematical approximations and manual feature screening processes, which have certain subjectivity and one-sidedness, limiting the performance of the model. In order to alleviate the above challenges, this paper proposes a machine learning (ML) model based on neural networks. The model can assist both PSCs design and analysis of their potential mechanism, demonstrating enhanced and comprehensive auxiliary capabilities. To make the model have higher feasibility and fit the real experimental process more closely, this paper collects the corresponding real experimental data from numerous research papers to develop the model. Compared with other classical ML methods, the proposed model achieved better overall performance. Regarding analysis of underlying mechanism, the relevant laws explored by the model are consistent with the actual experiment results of existing articles. The model exhibits great potential to discover complex laws that are difficult for humans to discover directly. In addition, we also fabricated PSCs to verify the guidance ability of the model in this paper for real experiments. Eventually, the model achieved acceptable results. This work provides new insights into integrating ML methods and PSC design techniques, as well as bridging photovoltaic power generation technology and other fields.

Integrating remote sensing and machine learning to detect turbidity anomalies in hydroelectric reservoirs

Monitoring water quality in reservoirs is essential for the maintenance of aquatic ecosystems and socioeconomic services. In this scenario, the observation of abrupt elevations of physicochemical parameters, such as turbidity and other indicators, can signal anomalies associated with the occurrence of critical events, requiring operational actions and planning to mitigate negative environmental impacts on water resources. This work aims to integrate Machine Learning methods specialized in anomaly detection with data obtained from remote sensing images to identify with high turbidity events in the surface water of the Três Marias Hydroelectric Reservoir. Four distinct threshold-based scenarios were evaluated, in which the overall performance, based on F1-score, showed decreasing trends as the thresholds became more restrictive. In general, the anomaly identification maps generated through the models ratified the applicability of the methods in the diagnosis of surface water in reservoirs in distinct hydrological contexts (dry and wet), effectively identifying locations with anomalous turbidity values.

Super learner approach to predict total organic carbon using stacking machine learning models based on well logs

Determining the total organic carbon (TOC) content is essential information for risk assessment in oil exploration, as it is a parameter used for the characterization of hydrocarbon-generating rocks, considering that intervals rich in organic matter are the basic requirements for oil and gas accumulation. However, the determination of TOC can be costly, demanding destructive tests in samples from the source rock, expensive laboratory machinery, and specialized personnel. In this context, one notes the necessity of the computational methods to bypass those problems and that machine learning models emerge as an option. One approach to integrating machine learning methods improves performance and, consequently, the prediction quality is stacking models. This paper presents a super learner strategy, based on stacking approaches, as a surrogate model for TOC modeling. The super-learner has three levels in this structure containing different types of learners (machine learning methods), where two stack models from the first two levels. The following machine learning models were used in the building of super learner the K-Neighbors Nearest (KNN), Linear Regression (LR), Multi-layer Perceptron Neural Network (MLP), Random Forest (RF), Ridge Regression (RR), and Support Vector Regression (SVR). The proposed model was compared with standalone machine learning models and other canonical stacking models. The resulting super learner stacking model attained the best average performance for the TOC modeling (R = 0.897, R2 = 0.80, RMSE = 1.16, MAE = 0.93, and MAPE = 28.30%). The proposed approach produces an alternative data-driven efficient model for TOC prediction, resulting in reliable automated technology to assist oil and gas well management and decision-making.

AREA-AFFINITY: A Web Server for Machine Learning-Based Prediction of Protein-Protein and Antibody-Protein Antigen Binding Affinities.

Protein-Protein binding affinity reflects the binding strength between the binding partners. The prediction of protein-protein binding affinity is important for elucidating protein functions and also for designing protein-based therapeutics. The geometric characteristics such as area (both interface and surface areas) in the structure of a protein-protein complex play an important role in determining protein-protein interactions and their binding affinity. Here, we present a free web server for academic use, AREA-AFFINITY, for prediction of protein-protein or antibody-protein antigen binding affinity based on interface and surface areas in the structure of a protein-protein complex. AREA-AFFINITY implements 60 effective area-based protein-protein affinity predictive models and 37 effective area-based models specific for antibody-protein antigen binding affinity prediction developed in our recent studies. These models take into consideration the roles of interface and surface areas in binding affinity by using areas classified according to different amino acid types with different biophysical nature. The models with the best performances integrate machine learning methods such as neural network or random forest. These newly developed models have superior or comparable performance compared to the commonly used existing methods. AREA-AFFINITY is available for free at: https://affinity.cuhk.edu.cn/.

Network modeling helps to tackle the complexity of drug-disease systems.

From the (patho)physiological point of view, diseases can be considered as emergent properties of living systems stemming from the complexity of these systems. Complex systems display some typical features, including the presence of emergent behavior and the organization in successive hierarchic levels. Drug treatments increase this complexity scenario, and from some years the use of network models has been introduced to describe drug-disease systems and to make predictions about them with regard to several aspects related to drug discovery. Here, we review some recent examples thereof with the aim to illustrate how network science tools can be very effective in addressing both tasks. We will examine the use of bipartite networks that lead to the important concept of "disease module", as well as the introduction of more articulated models, like multi-scale and multiplex networks, able to describe disease systems at increasing levels of organization. Examples of predictive models will then be discussed, considering both those that exploit approaches purely based on graph theory and those that integrate machine learning methods. A short account of both kinds of methodological applications will be provided. Finally, the point will be made on the present situation of modeling complex drug-disease systems highlighting some open issues. This article is categorized under: Neurological Diseases > Computational Models Infectious Diseases > Computational Models Cardiovascular Diseases > Computational Models.

The use of data independent acquisition based proteomic analysis and machine learning to reveal potential biomarkers for autism spectrum disorder

Autism spectrum disorder (ASD) is a complex neurological developmental disorder in children, and is associated with social isolation and restricted interests. The etiology of this disorder is still unknown. There is neither any confirmed laboratory test nor any effective therapeutic strategy to diagnose or cure it. We performed data independent acquisition (DIA) and multiple reaction monitoring (MRM) analysis of plasma from children with ASD and controls. The result showed that 45 differentially expressed proteins (DEPs) were identified between autistic subjects and controls. Among these, only one DEP was down-regulated in ASD; other DEPs were up-regulated in ASD children's plasma. These proteins are found associated with complement and coagulation cascades, vitamin digestion and absorption, cholesterol metabolism, platelet degranulation, selenium micronutrient network, extracellular matrix organization and inflammatory pathway, which have been reported to be related to ASD. After MRM verification, five key proteins in complement pathway (PLG, SERPINC1, and A2M) and inflammatory pathway (CD5L, ATRN, SERPINC1, and A2M) were confirmed to be significantly up-regulated in ASD group. Through the screening of machine learning model and MRM verification, we found that two proteins (biotinidase and carbonic anhydrase 1) can be used as early diagnostic markers of ASD (AUC = 0.8, p = 0.0001). SignificanceASD is the fastest growing neurodevelopmental disorder in the world and has become a major public health problem worldwide. Its prevalence has been steadily increasing, with a global prevalence rate of 1%. Early diagnosis and intervention can achieve better prognosis. In this study, data independent acquisition (DIA) and multiple reaction monitoring (MRM) analysis was applied to analyze the plasma proteome of ASD patients (31 (±5) months old), and 378 proteins were quantified. 45 differentially expressed proteins (DEPs) were identified between the ASD group and the control group. They mainly were associated with platelet degranulation, ECM proteoglycar, complement and coagulation cascades, selenium micronutrient network, regulation of insulin-like growth factor (IGF) transport and uptake by insulin-like growth factor binding proteins (IGFBPs), cholesterol metabolism, vitamin metabolism, and inflammatory pathway. Through the integrated machine learning methods and the MRM verification of independent samples, it is considered that biotinidase and carbon anhydrase 1 have the potential to become biomarkers for the early diagnosis of ASD. These results complement proteomics database of the ASD patients, broaden our understanding of ASD, and provide a panel of biomarkers for the early diagnosis of ASD.

Handling Class Imbalance and Class Overlap in Machine Learning Applications for Undeclared Work Prediction

Undeclared work is a composite socioeconomic matter severely affecting the welfare of workers, legitimate companies, and the state by issuing unfair competition in the labour market and causing considerable state revenue losses by tax evasion. Labour inspectorates are tasked to deal effectively with this issue but usually lack adequate resources and proper tools, yet they own large volumes of past inspection data that, if aptly processed through innovative machine learning techniques, may produce understandable insights into the extent and prevailing patterns of undeclared work and efficient tools to address it. Such datasets are typically imbalanced regarding undeclared work, and contain overlapping inspection discoveries, two issues that impede the learning process. This research points to the problems of class imbalance and class overlap in this domain and applies combinations of data engineering techniques to address them using a dataset of 16.7 K actual labour inspections. Three associative classification algorithms are employed, and multiple classifiers are built and assessed for their predictability and interpretability. The study indicates the overall benefits for the inspection authorities when integrating machine learning methods in targeting undeclared work and proves considerable prediction performance improvement when following data engineering approaches to address the class imbalance and class overlap issues.

Integrated Machine Learning Decision Tree Model for Risk Evaluation in Patients with Non-Valvular Atrial Fibrillation When Taking Different Doses of Dabigatran.

The new generation of nonvitamin K antagonists are broadly applied for stroke prevention due to their notable efficacy and safety. Our study aimed to develop a suggestive utilization of dabigatran through an integrated machine learning (ML) decision-tree model. Participants taking different doses of dabigatran in the Randomized Evaluation of Long-Term Anticoagulant Therapy trial were included in our analysis and defined as the 110 mg and 150 mg groups. The proposed scheme integrated ML methods, namely naive Bayes, random forest (RF), classification and regression tree (CART), and extreme gradient boosting (XGBoost), which were used to identify the essential variables for predicting vascular events in the 110 mg group and bleeding in the 150 mg group. RF (0.764 for 110 mg; 0.747 for 150 mg) and XGBoost (0.708 for 110 mg; 0.761 for 150 mg) had better area under the receiver operating characteristic curve (AUC) values than logistic regression (benchmark model; 0.683 for 110 mg; 0.739 for 150 mg). We then selected the top ten important variables as internal nodes of the CART decision tree. The two best CART models with ten important variables output tree-shaped rules for predicting vascular events in the 110 mg group and bleeding in the 150 mg group. Our model can be used to provide more visualized and interpretable suggestive rules to clinicians managing NVAF patients who are taking dabigatran.

IMPROVINGURBAN ENERGYRESILIENCEWITH AN INTEGRATIVE FRAMEWORK BASED ON MACHINE LEARNINGMETHODS

Introduction:Climate change and global warming are among the greatest challenges facing the world today. A newconcept, known as urban resilience, has been developed in response. There are various approaches to urban resilience. Among them, is the urban energy resilience (UER) approach, which poses a considerable challenge. Machine learning (ML), as an application of artificial intelligence (AI), provides powerful and affordable computing resources, large-scale datamining, advanced algorithms, and real-timemonitoring. However, veryfewstudies have investigated howsuch aspectscan be integrated into urban resilience in general, and UER in particular. Purpose of the study: The study develops an integrative framework that can improve UER, based on ML methods. Methodology: We carried out a bibliometric analysisand a systematic review of UER in accordance with AI concepts, models, and applications. Results:The findings of this study were used to create an integrative framework, based on three hierarchical phases, which effectively addressed the main capabilities of UER, identified its priorities, and shed light on how ML can benefit UER as a whole. Novelty:The framework developed in this study also offers insights in integrating ML methods into UER as strategically as possible, especially in the context of climate change and urban energy systems. This framework can serve as reference for specialistsand decision-makers aiming to expand AI and ML applications to optimize UER.

Application of Machine Learning in Spatial Proteomics.

Spatial proteomics is an interdisciplinary field that investigates the localization and dynamics of proteins, and it has gained extensive attention in recent years, especially the subcellular proteomics. Numerous evidence indicate that the subcellular localization of proteins is associated with various cellular processes and disease progression. Mass spectrometry (MS)-based and imaging-based experimental approaches have been developed to acquire large-scale spatial proteomic data. To allow the reliable analysis of increasingly complex spatial proteomics data, machine learning (ML) methods have been widely used in both MS-based and imaging-based spatial proteomic data analysis pipelines. Here, we comprehensively survey the applications of ML in spatial proteomics from following aspects: (1) data resources for spatial proteome are comprehensively introduced; (2) the roles of different ML algorithms in data analysis pipelines are elaborated; (3) successful applications of spatial proteomics and several analytical tools integrating ML methods are presented; (4) challenges existing in modern ML-based spatial proteomics studies are discussed. This review provides guidelines for researchers seeking to apply ML methods to analyze spatial proteomic data and can facilitate insightful understanding of cell biology as well as the future research in medical and drug discovery communities.

Integrated machine learning methods identify FNDC3B as a potential prognostic biomarker and correlated with immune infiltrates in glioma.

BackgroundRecent discoveries have revealed that fibronectin type III domain containing 3B (FNDC3B) acts as an oncogene in various cancers; however, its role in glioma remains unclear.MethodsIn this study, we comprehensively investigated the expression, prognostic value, and immune significance of FNDC3B in glioma using several databases and a variety of machine learning algorithms. RNA expression data and clinical information of 529 patients from the Cancer Genome Atlas (TCGA) and 1319 patients from Chinese Glioma Genome Atlas (CGGA) databases were downloaded for further investigation. To evaluate whether FNDC3B expression can predict clinical prognosis of glioma, we constructed a clinical nomogram to estimate long-term survival probabilities. The predicted nomogram was validated by CGGA cohorts. Differentially expressed genes (DEGs) were detected by the Wilcoxon test based on the TCGA-LGG dataset and the weighted gene co-expression network analysis (WGCNA) was implemented to identify the significant module associated with the expression level of FNDC3B. Furthermore, we investigated the correlation between FNDC3B with cancer immune infiltrates using TISIDB, ESTIMATE, and CIBERSORTx.ResultsHigher FNDC3B expression displayed a remarkably worse overall survival and the expression level of FNDC3B was an independent prognostic indicator for patients with glioma. Based on TCGA LGG dataset, a co-expression network was established and the hub genes were identified. FNDC3B expression was positively correlated to the tumor-infiltrating lymphocytes and immune infiltration score, and high FNDC3B expression was accompanied by the increased expression of B7-H3, PD-L1, TIM-3, PD-1, and CTLA-4. Moreover, expression of FNDC3B was significantly associated with infiltrating levels of several types of immune cells and most of their gene markers in glioma.ConclusionThis study demonstrated that FNDC3B may be involved in the occurrence and development of glioma and can be regarded as a promising prognostic and immunotherapeutic biomarker for the treatment of glioma.

RMHIL: A Rule Matching Algorithm Based on Heterogeneous Integrated Learning in Software Defined Network.

To ensure the efficient operation of large-scale networks, the flow scheduling in the software defined network (SDN) requires the matching time and memory overhead of rule matching to be as low as possible. To meet the requirement, we solve the rule matching problem by integrating machine learning methods, including recurrent neural networks, reinforcement learning, and decision trees. We first describe the SDN rule matching problem and transform it into a heterogeneous integrated learning problem. Then, we design and implement an SDN flow forwarding rule matching algorithm based on heterogeneous integrated learning, referred to as RMHIL. Finally, we compare RMHIL with two existing algorithms, and the comparative experimental results show that RMHIL has advantages in matching time and memory overhead.