Steel Surface Defect Recognition in Smart Manufacturing Using Deep Ensemble Transfer Learning-Based Techniques

Smart manufacturing and Industry 4.0 are transforming traditional manufacturing processes by utilizing innovative technologies such as the artificial intelligence(AI)and internet of things(IoT)to enhance efficiency,reduce costs,and ensure product quality.In light of the recent advancement of Industry 4.0,identifying defects has become important for ensuring the quality of products during the manufacturing process.In this research,we present an ensemble methodology for accurately classifying hot rolled steel surface defects by combining the strengths of four pre-trained convolutional neural network(CNN)architectures:VGG16,VGG19,Xception,and Mobile-Net V2,compensating for their individual weaknesses.We evaluated our methodology on the Xsteel surface defect dataset(XSDD),which comprises seven different classes.The ensemble methodology integrated the predictions of individual models through two methods:model averaging and weighted averaging.Our evaluation showed that the model averaging ensemble achieved an accuracy of 98.89%,a recall of 98.92%,a precision of 99.05%,and an F1-score of 98.97%,while the weighted averaging ensemble reached an accuracy of 99.72%,a recall of 99.74%,a precision of 99.67%,and an F1-score of 99.70%.The proposed weighted averaging ensemble model outperformed the model averaging method and the individual models in detecting defects in terms of accuracy,recall,precision,and F1-score.Comparative analysis with recent studies also showed the superior performance of our methodology.

Landslide Susceptibility Mapping Using RBFN-Based Ensemble Machine Learning Models

This study was aimed to prepare landslide susceptibility maps for the Pithoragarh district in Uttarakhand,India,using advanced ensemble models that combined Radial Basis Function Networks(RBFN)with three ensemble learning techniques:DAGGING(DG),MULTIBOOST(MB),and ADABOOST(AB).This combination resulted in three distinct ensemble models:DG-RBFN,MB-RBFN,and AB-RBFN.Additionally,a traditional weighted method,Information Value(IV),and a benchmark machine learning(ML)model,Multilayer Perceptron Neural Network(MLP),were employed for comparison and validation.The models were developed using ten landslide conditioning factors,which included slope,aspect,elevation,curvature,land cover,geomorphology,overburden depth,lithology,distance to rivers and distance to roads.These factors were instrumental in predicting the output variable,which was the probability of landslide occurrence.Statistical analysis of the models’performance indicated that the DG-RBFN model,with an Area Under ROC Curve(AUC)of 0.931,outperformed the other models.The AB-RBFN model achieved an AUC of 0.929,the MB-RBFN model had an AUC of 0.913,and the MLP model recorded an AUC of 0.926.These results suggest that the advanced ensemble ML model DG-RBFN was more accurate than traditional statistical model,single MLP model,and other ensemble models in preparing trustworthy landslide susceptibility maps,thereby enhancing land use planning and decision-making.

基于改进SMOTE算法和Ensemble模型的学习结果预测方法

为解决不同领域的数据分类和预测任务中单个机器学习算法适用性较差的问题,以及缓解数据集严重不平衡对预测性能的影响,提出了基于合成少数类过采样(SMOTE)和Ensemble集成模型的数据分类方法。传统SMOTE算法通过对少数类样本进行插值来生成新的合成样本,合成样本中存在噪声和样本间相似性较高的问题。为此,提出了改进的SMOTE算法,通过距离计算移除噪声样本和易混淆样本,得到高区分度的纯净合成样本。然后,利用Ensemble方法调整样本和分类器权重,并组成分类效果更好的强分类器。在公开在线学习数据集Kalboard360上的实验结果表明,使用极限随机树(ERT)分类器时,结合改进SMOTE和Ensemble模型后实现了97.9%的预测准确度,比单个ERT分类器提升了5.5%,证明所提改进SMOTE算法能够生成高质量的均衡化数据,且集成学习模型的性能显著优于单个机器学习算法。

Selective ensemble modeling based on nonlinear frequency spectral feature extraction for predicting load parameter in ball mills

Strong mechanical vibration and acoustical signals of grinding process contain useful information related to load parameters in ball mills. It is a challenge to extract latent features and construct soft sensor model with high dimensional frequency spectra of these signals. This paper aims to develop a selective ensemble modeling approach based on nonlinear latent frequency spectral feature extraction for accurate measurement of material to ball volume ratio. Latent features are first extracted from different vibrations and acoustic spectral segments by kernel partial least squares. Algorithms of bootstrap and least squares support vector machines are employed to produce candidate sub-models using these latent features as inputs. Ensemble sub-models are selected based on genetic algorithm optimization toolbox. Partial least squares regression is used to combine these sub-models to eliminate collinearity among their prediction outputs. Results indicate that the proposed modeling approach has better prediction performance than previous ones.

Prediction of malignancy selective neural networks degree in brain glioma using ensemble

A clustering algorithm based selective neural networks ensemble （CLUSEN） is proposed to predict the degree of malignancy in brain glioma. Since the degree prediction of malignancy is critical before brain surgery, many learning methods are used like rule induction algorithm, single neural networks, support vector machines, etc. Ensemble learning methods can improve the generalization of single learning machine, and are becoming popular in the machine learning and medical data processing communities. The procedure of CLUSEN can efficiently remove redundancy learning individuals and help improve the diversity of ensemble methods. CLUSEN is used to predict the degree of malignancy in brain glioma. Experimental results on a set of brain glioma data show that, compared to support vector machines, rule induction and single neural networks, the classification accuracy of CLUSEN is higher.

Clustering-based selective neural network ensemble

An effective ensemble should consist of a set of networks that are both accurate and diverse. We propose a novel clustering-based selective algorithm for constructing neural network ensemble, where clustering technology is used to classify trained networks according to similarity and optimally select the most accurate individual network from each cluster to make up the ensemble. Empirical studies on regression of four typical datasets showed that this approach yields significantly smaller en- semble achieving better performance than other traditional ones such as Bagging and Boosting. The bias variance decomposition of the predictive error shows that the success of the proposed approach may lie in its properly tuning the bias/variance trade-off to reduce the prediction error (the sum of bias2 and variance).

A New Selective Neural Network Ensemble Method Based on Error Vectorization and Its Application in High-density Polyethylene (HDPE) Cascade Reaction Process

Chemical processes are complex, for which traditional neural network models usually can not lead to satisfactory accuracy. Selective neural network ensemble is an effective way to enhance the generalization accuracy of networks, but there are some problems, e.g., lacking of unified definition of diversity among component neural networks and difficult to improve the accuracy by selecting if the diversities of available networks are small. In this study, the output errors of networks are vectorized, the diversity of networks is defined based on the error vectors, and the size of ensemble is analyzed. Then an error vectorization based selective neural network ensemble (EVSNE) is proposed, in which the error vector of each network can offset that of the other networks by training the component networks orderly. Thus the component networks have large diversity. Experiments and comparisons over standard data sets and actual chemical process data set for production of high-density polyethylene demonstrate that EVSNE performs better in generalization ability.

Prediction of users online purchase behavior based on selective ensemble learning

A probabilistic multi-dimensional selective ensemble learning method and its application in the prediction of users' online purchase behavior are studied in this work.Firstly, the classifier is integrated based on the dimension of predicted probability, and the pruning algorithm based on greedy forward search is obtained by combining the two indicators of accuracy and complementarity.Then the pruning algorithm is integrated into the Stacking ensemble method to establish a user online shopping behavior prediction model based on the probabilistic multi-dimensional selective ensemble method.Finally, the research method is compared with the prediction results of individual learners in ensemble learning and the Stacking ensemble method without pruning.The experimental results show that the proposed method can reduce the scale of integration, improve the prediction accuracy of the model, and predict the user's online purchase behavior.

EDSUCh:A robust ensemble data summarization method for effective medical diagnosis

Identifying rare patterns for medical diagnosis is a challenging task due to heterogeneity and the volume of data.Data summarization can create a concise version of the original data that can be used for effective diagnosis.In this paper,we propose an ensemble summarization method that combines clustering and sampling to create a summary of the original data to ensure the inclusion of rare patterns.To the best of our knowledge,there has been no such technique available to augment the performance of anomaly detection techniques and simultaneously increase the efficiency of medical diagnosis.The performance of popular anomaly detection algorithms increases significantly in terms of accuracy and computational complexity when the summaries are used.Therefore,the medical diagnosis becomes more effective,and our experimental results reflect that the combination of the proposed summarization scheme and all underlying algorithms used in this paper outperforms the most popular anomaly detection techniques.

Advances in selective conversion of carbohydrates into 5-hydroxymethylfurfural

Converting carbohydrates into 5-hydroxymethylfurfural(5-HMF) is an attractive and promising route for value-added utilization of agricultural and forestry biomass resource. As an important platform compound, 5-HMF possesses high active furan structure with hydroxymethyl and aldehyde group for production of various bio-chemicals and materials, meanwhile, which suffer from low stability and poor yield during the industrial biorefinery process. Hence, selective production of 5-HMF with high-yield and low-cost has attracted extensive attention from scientific and industrial researchers. This review sorted and described the latest advanced research on solvent and catalyst system, as well as energy field effect for production of 5-HMF with different feedstock in detail, emphatically discussing the solvent effect and its synergistic effect with other aspects. Besides, the future prospects and challenges for production of 5-HMF from carbohydrates were also presented, which provide a profound insight into industrial 5-HMF process with economic and environmental feature.

Dealuminated Hβ zeolite for selective conversion of fructose to furfural and formic acid

The fructose-to-furfural transformation is facing major challenges in the selectivity and high efficiency. Herein, we have developed a simple and effective approach for the selective conversion of fructose to furfural using Hβ zeolite modified by organic acids for dealuminization to regulate its textural and acidic properties. It was found that citric acid-dealuminized Hβ zeolite possessed high specific surface areas, wide channels and high Brønsted acid amount, which facilitated the selective conversion of fructose to furfural with a maximum yield of 76.2% at433 K for 1 h in the γ-butyrolactone(GBL)-H_(2)O system, as well as the concomitant formation of 83.0% formic acid. The^(13)C-isotope labelling experiments and the mechanism revealed that the selective cleavage of C1–C2 or C5–C6 bond on fructose was firstly occurred to form pentose or C5 intermediate by weak Brønsted acid, which was then dehydrated to furfural by strong Brønsted acid. Also this dealuminized Hβ catalyst showed the great recycling performance and was active for the conversion of glucose and mannose.

Synergy of heterogeneous Co/Ni dual atoms enabling selective C-O bond scission of lignin coupling with in-situ N-functionalization

Selective cleavage of Csp^(2)-OCH_(3)bond in lignin without breaking other types of C-O bonds followed by N-functionalization is fascinating for on-purpose valorization of biomass.Here,a Co/Ni-based dual-atom catalyst CoNiDA@NC prepared by in-situ evaporation and acid-etching of metal species from tailor-made metal–organic frameworks was efficient for reductive upgrading of various lignin-derived phenols to cyclohexanols(88.5%–99.9%yields),which had ca.4 times higher reaction rate than the single-atom catalyst and was superior to state-of-the-art heterogeneous catalysts.The synergistic catalysis of Co/Ni dual atoms facilitated both hydrogen dissociation and hydrogenolysis steps,and could optimize adsorption configuration of lignin-derived methoxylated phenols to further favor the Csp^(2)-OCH_(3)cleavage,as elaborated by theoretical calculations.Notably,the CoNi_(DA)@NC catalyst was highly recyclable,and exhibited excellent demethoxylation performance(77.1%yield)in real lignin monomer mixtures.Via in-situ cascade conversion processes assisted by dual-atom catalysis,various high-value N-containing chemicals,including caprolactams and cyclohexylamines,could be produced from lignin.

Non-crossing Quantile Regression Neural Network as a Calibration Tool for Ensemble Weather Forecasts

Despite the maturity of ensemble numerical weather prediction(NWP),the resulting forecasts are still,more often than not,under-dispersed.As such,forecast calibration tools have become popular.Among those tools,quantile regression(QR)is highly competitive in terms of both flexibility and predictive performance.Nevertheless,a long-standing problem of QR is quantile crossing,which greatly limits the interpretability of QR-calibrated forecasts.On this point,this study proposes a non-crossing quantile regression neural network(NCQRNN),for calibrating ensemble NWP forecasts into a set of reliable quantile forecasts without crossing.The overarching design principle of NCQRNN is to add on top of the conventional QRNN structure another hidden layer,which imposes a non-decreasing mapping between the combined output from nodes of the last hidden layer to the nodes of the output layer,through a triangular weight matrix with positive entries.The empirical part of the work considers a solar irradiance case study,in which four years of ensemble irradiance forecasts at seven locations,issued by the European Centre for Medium-Range Weather Forecasts,are calibrated via NCQRNN,as well as via an eclectic mix of benchmarking models,ranging from the naïve climatology to the state-of-the-art deep-learning and other non-crossing models.Formal and stringent forecast verification suggests that the forecasts post-processed via NCQRNN attain the maximum sharpness subject to calibration,amongst all competitors.Furthermore,the proposed conception to resolve quantile crossing is remarkably simple yet general,and thus has broad applicability as it can be integrated with many shallow-and deep-learning-based neural networks.

Selective leaching of lithium from spent lithium-ion batteries using sulfuric acid and oxalic acid

Traditional hydrometallurgical methods for recovering spent lithium-ion batteries(LIBs)involve acid leaching to simultaneously extract all valuable metals into the leachate.These methods usually are followed by a series of separation steps such as precipitation,extraction,and stripping to separate the individual valuable metals.In this study,we present a process for selectively leaching lithium through the synergistic effect of sulfuric and oxalic acids.Under optimal leaching conditions(leaching time of 1.5 h,leaching temperature of 70°C,liquid-solid ratio of 4 mL/g,oxalic acid ratio of 1.3,and sulfuric acid ratio of 1.3),the lithium leaching efficiency reached89.6%,and the leaching efficiencies of Ni,Co,and Mn were 12.8%,6.5%,and 21.7%.X-ray diffraction(XRD)and inductively coupled plasma optical emission spectrometer(ICP-OES)analyses showed that most of the Ni,Co,and Mn in the raw material remained as solid residue oxides and oxalates.This study offers a new approach to enriching the relevant theory for selectively recovering lithium from spent LIBs.

On the role of cellular microstructure in austenite reversion in selective laser melted maraging steel

Cellular microstructure is a unique feature in alloys fabricated by selective laser melting(SLM).Abundant efforts have been made to reveal the formation mechanism of cellular microstructures and its influences on mechanical performances,while its potential role in microstructure architecting during post-heat treatment is rarely explored.In this work,we investigated the features of cellular microstructures in an SLM-fabricated 18Ni(300)steel and revealed how this microstructure influences austenite reversion upon aging.Segregation of Ti and Mo is experimentally detected at cell boundaries.It is interestingly found that a distinctive reverted austenite network forms rapidly along cell boundaries during aging,whereas much less austenite is found in conventionally treated 18Ni(300)steels.The rapid austenite reversion in SLM-fabricated material proceeds mainly via the growth of retained austenite on cell boundaries while the nucleation and growth of new austenite grains is negligible.Phase-field simulations suggest austenite grows in a fast,partitionless manner along cell boundaries where the chemical driving force for austen-ite reversion is substantially enhanced by Ti and Mo segregations,but in a sluggish,partitioning manner towards cell interiors.Contrary to conventional views that austenite fraction should be confined to avoid strength reduction,current SLM-fabricated 18Ni(300)steel containing∼13%cellular austenite is found to have higher tensile strength compared to its counterparts with negligible austenite.The design of austen-ite also shows its potential to enhance fracture toughness.The current study demonstrates that cellular structures could substantially alter austenite reversion behavior,providing a new route for microstructure architecting in additively manufactured steels.

Detection and defending the XSS attack using novel hybrid stacking ensemble learning-based DNN approach

Existing web-based security applications have failed in many situations due to the great intelligence of attackers.Among web applications,Cross-Site Scripting(XSS)is one of the dangerous assaults experienced while modifying an organization's or user's information.To avoid these security challenges,this article proposes a novel,all-encompassing combination of machine learning(NB,SVM,k-NN)and deep learning(RNN,CNN,LSTM)frameworks for detecting and defending against XSS attacks with high accuracy and efficiency.Based on the representation,a novel idea for merging stacking ensemble with web applications,termed“hybrid stacking”,is proposed.In order to implement the aforementioned methods,four distinct datasets,each of which contains both safe and unsafe content,are considered.The hybrid detection method can adaptively identify the attacks from the URL,and the defense mechanism inherits the advantages of URL encoding with dictionary-based mapping to improve prediction accuracy,accelerate the training process,and effectively remove the unsafe JScript/JavaScript keywords from the URL.The simulation results show that the proposed hybrid model is more efficient than the existing detection methods.It produces more than 99.5%accurate XSS attack classification results(accuracy,precision,recall,f1_score,and Receiver Operating Characteristic(ROC))and is highly resistant to XSS attacks.In order to ensure the security of the server's information,the proposed hybrid approach is demonstrated in a real-time environment.

Recent developments in selective laser processes for wearable devices

Recently,the increasing interest in wearable technology for personal healthcare and smart virtual/augmented reality applications has led to the development of facile fabrication methods.Lasers have long been used to develop original solutions to such challenging technological problems due to their remote,sterile,rapid,and site-selective processing of materials.In this review,recent developments in relevant laser processes are summarized under two separate categories.First,transformative approaches,such as for laser-induced graphene,are introduced.In addition to design optimization and the alteration of a native substrate,the latest advances under a transformative approach now enable more complex material compositions and multilayer device configurations through the simultaneous transformation of heterogeneous precursors,or the sequential addition of functional layers coupled with other electronic elements.In addition,the more conventional laser techniques,such as ablation,sintering,and synthesis,can still be used to enhance the functionality of an entire system through the expansion of applicable materials and the adoption of new mechanisms.Later,various wearable device components developed through the corresponding laser processes are discussed,with an emphasis on chemical/physical sensors and energy devices.In addition,special attention is given to applications that use multiple laser sources or processes,which lay the foundation for the all-laser fabrication of wearable devices.

Selective internal radiation therapy segmentectomy:A new minimally invasive curative option for primary liver malignancies?

Transarterial radioembolization or selective internal radiation therapy(SIRT)has emerged as a minimally invasive approach for the treatment of tumors.This percutaneous technique involves the local,intra-arterial delivery of radioactive microspheres directly into the tumor.Historically employed as a palliative measure for liver malignancies,SIRT has gained traction over the past decade as a potential curative option,mirroring the increasing role of radiation segmentectomy.The latest update of the BCLC hepatocellular carcinoma guidelines recognizes SIRT as an effective treatment modality comparable to other local ablative methods,particularly well-suited for patients where surgical resection or ablation is not feasible.Radiation segmentectomy is a more selective approach,aiming to deliver high-dose radiation to one to three specific hepatic segments,while minimizing damage to surrounding healthy tissue.Future research efforts in radiation segmentectomy should prioritize optimizing radiation dosimetry and refining the technique for super-selective administration of radiospheres within the designated hepatic segments.

A New Speed Limit Recognition Methodology Based on Ensemble Learning:Hardware Validation

Advanced DriverAssistance Systems(ADAS)technologies can assist drivers or be part of automatic driving systems to support the driving process and improve the level of safety and comfort on the road.Traffic Sign Recognition System(TSRS)is one of themost important components ofADAS.Among the challengeswith TSRS is being able to recognize road signs with the highest accuracy and the shortest processing time.Accordingly,this paper introduces a new real time methodology recognizing Speed Limit Signs based on a trio of developed modules.Firstly,the Speed Limit Detection(SLD)module uses the Haar Cascade technique to generate a new SL detector in order to localize SL signs within captured frames.Secondly,the Speed Limit Classification(SLC)module,featuring machine learning classifiers alongside a newly developed model called DeepSL,harnesses the power of a CNN architecture to extract intricate features from speed limit sign images,ensuring efficient and precise recognition.In addition,a new Speed Limit Classifiers Fusion(SLCF)module has been developed by combining trained ML classifiers and the DeepSL model by using the Dempster-Shafer theory of belief functions and ensemble learning’s voting technique.Through rigorous software and hardware validation processes,the proposedmethodology has achieved highly significant F1 scores of 99.98%and 99.96%for DS theory and the votingmethod,respectively.Furthermore,a prototype encompassing all components demonstrates outstanding reliability and efficacy,with processing times of 150 ms for the Raspberry Pi board and 81.5 ms for the Nano Jetson board,marking a significant advancement in TSRS technology.

Identification Method for Users-Transformer Relationship in Station Area Based on Local Selective Combination in Parallel Outlier Ensembles Algorithm

In the power distribution system,the missing or incorrect file of users-transformer relationship(UTR)in lowvoltage station area(LVSA)will affect the leanmanagement of the LVSA,and the operation andmaintenance of the distribution network.To effectively improve the lean management of LVSA,the paper proposes an identification method for the UTR based on Local Selective Combination in ParallelOutlier Ensembles algorithm(LSCP).Firstly,the voltage data is reconstructed based on the information entropy to highlight the differences in between.Then,the LSCP algorithmcombines four base outlier detection algorithms,namely Isolation Forest(I-Forest),One-Class Support VectorMachine(OC-SVM),Copula-Based Outlier Detection(COPOD)and Local Outlier Factor(LOF),to construct the identification model of UTR.This model can accurately detect users’differences in voltage data,and identify users with wrong UTR.Meanwhile,the key input parameter of the LSCP algorithm is determined automatically through the line loss rate,and the influence of artificial settings on recognition accuracy can be reduced.Finally,thismethod is verified in the actual LVSA where the recall and precision rates are 100%compared with othermethods.Furthermore,the applicability to the LVSAs with difficult data acquisition and the voltage data error in transmission are analyzed.The proposed method adopts the ensemble learning framework and does not need to set the detection threshold manually.And it is applicable to the LVSAs with difficult data acquisition and high voltage similarity,which improves the stability and accuracy of UTR identification in LVSA.