Search

The work presents a newly designed penalty function to be added with an existing Cross Entropy based fitness function [3] for optimal selection of multi-level thresholds for image segmentation. The extended fitness function so designed is tested here for Brain magnetic resonance (MR) image segmentation using nature-inspired meta-heuristics such as Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Whale Optimization Algorithm (WOA) and Moth-flame Optimization (MFO), and MFO is finally selected. The proposed method excels in Brain MR image segmentation when tested on WBA database and BrainWeb MR image database with other nature-inspired meta-heuritic based methods. It outperformed the other methods in terms of the three performance metrics Peak Signal to Noise Ratio (PSNR), Structural Similarity Index (SSIM) and Feature Similarity Index (FSIM). The best results shown by the proposed method on WBA database in terms of these metrics - PSNR, SSIM and FSIM - are 29.77, 0.927 and 0.899 respectively. On BrainWeb database, the proposed method yields 26.91, 0.864 and 0.892 for the three respective metrics. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Chongqing University of Posts & Telecommunications (Natural Science Edition) is the property of Chongqing University of Posts & Telecommunications and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

With a rapid development of artificial intelligence technology, fine-grained image classification has gained widespread application. For mobile terminals, this paper introduces an image classification method built on MobileViT, and it can apply into fine-grained image classification. The original MobileViT model has been optimized in three ways. Initially, the h-swish activation function is used to enhance the network performance. Second, the cross-entropy loss function is used to further realize the parameter optimization and model accuracy improvement. Finally, a dropout layer is joined before the fully connected layer can effectively decrease the model recognition time and prevent over-fitting. Experimental data on public tomato disease datasets demonstrate that the improved fine-grained image classification method put forward in this paper exhibits higher classification accuracy, better stability and network generalization ability than other models. [ABSTRACT FROM AUTHOR]

Medical imaging description and disease diagnosis are vitally important yet time-consuming. Automated diagnosis report generation (DRG) from medical imaging description can reduce clinicians' workload and improve their routine efficiency. To address this natural language generation task, fine-tuning a pre-trained large language model (LLM) is cost-effective and indispensable, and its success has been witnessed in many downstream applications. However, semantic inconsistency of sentence embeddings has been massively observed from undesirable repetitions or unnaturalness in text generation. To address the underlying issue of anisotropic distribution of token representation, in this study, a contrastive learning penalized cross-entropy (CLpCE) objective function is implemented to enhance the semantic consistency and accuracy of token representation by guiding the fine-tuning procedure towards a specific task. Furthermore, to improve the diversity of token generation in text summarization and to prevent sampling from unreliable tail of token distributions, a diversity contrastive search (DCS) decoding method is designed for restricting the report generation derived from a probable candidate set with maintained semantic coherence. Furthermore, a novel metric named the maximum of token repetition ratio (maxTRR) is proposed to estimate the token diversity and to help determine the candidate output. Based on the LLM of a generative pre-trained Transformer 2 (GPT-2) of Chinese version, the proposed CLpCE with DCS (CLpCEwDCS) decoding framework is validated on 30,000 desensitized text samples from the "Medical Imaging Diagnosis Report Generation" track of 2023 Global Artificial Intelligence Technology Innovation Competition. Using four kinds of metrics evaluated from n-gram word matching, semantic relevance, and content similarity as well as the maxTRR metric extensive experiments reveal that the proposed framework effectively maintains semantic coherence and accuracy (BLEU-1, 0.4937; BLEU-2, 0.4107; BLEU-3, 0.3461; BLEU-4, 0.2933; METEOR, 0.2612; ROUGE, 0.5182; CIDER, 1.4339) and improves text generation diversity and naturalness (maxTRR, 0.12). The phenomenon of dull or repetitive text generation is common when fine-tuning pre-trained LLMs for natural language processing applications. This study might shed some light on relieving this issue by developing comprehensive strategies to enhance semantic coherence, accuracy and diversity of sentence embeddings. [ABSTRACT FROM AUTHOR]

In recent years, the lack of coordination in cross-border logistics has been one of the challenges and challenges faced by cross-border e-commerce. As the primary link in cross-border logistics, the selection of logistics service providers is an important foundation for promoting the development of cross-border e-commerce, and also a key link in improving the competitiveness of cross-border e-commerce enterprises. How to choose suitable and effective cross-border e-commerce logistics service providers has important theoretical significance and practical application value. The cross-border e-commerce logistics service providers evaluation is a multiple-attributed decision-making (MADM) problem. In this paper, the Type-2 neutrosophic number cross-entropy (T2NN-CE) technique is designed with help of cross-entropy and Type-2 neutrosophic number (T2NN). Furthermore, Then, T2NN-CE technique is built to solve the MADM. Finally, a numerical example for cross-border e-commerce logistics service providers evaluation is given and some comparisons are conducted to illustrate advantages of the designed T2NN-CE technique. The research contribution of the paper is outlined: (1) The T2NN-CE is managed under T2NNs; (2) the T2NN-CE method is implemented for MADM under T2NNs; (3) the T2NN-CE technique for cross-border e-commerce logistics service providers evaluation is constructed and were compared with some existing techniques; (4) Through the comparison, it is known that T2NN-CE technique for cross-border e-commerce logistics service providers evaluation is effective. [ABSTRACT FROM AUTHOR]

The T-spherical fuzzy set (T-SFS) has been widely used in multi-attribute decision-making (MADM) problems due to its powerful expression ability for uncertain information. The evidential reasoning (ER) method can avoid the information loss when aggregating the evaluation values of an alternative with regard to all attributes, which can prevent the decision result from being dominated by extreme evaluation values. Considering the significant advantage of the ER method in information aggregation, this paper develops a new aggregation model assimilating the T-SFSs and the ER method, which is called evidential T-spherical fuzzy MADM (ET-SFMADM). The ET-SFMADM can prevent information loss in the decision process, so as to obtain a correct ranking of the alternatives. Furthermore, this paper also proposes a T-spherical fuzzy cross entropy (T-SFCE) to determine the weights of attributes, which can reduce dependence on decision-makers. Finally, the viability of the ET-SFMADM is illustrated through a numerical example for the implementation effect of the hierarchical medical treatment system in different regions, and further the sensitivity analysis and comparison analysis with existing MADM methods are carried out to demonstrate the advantages of the proposed method. [ABSTRACT FROM AUTHOR]

In deep learning of classifiers, the cost function usually takes the form of a combination of SoftMax and CrossEntropy functions. The SoftMax unit transforms the scores predicted by the model network into assessments of the degree (probabilities) of an object's membership to a given class. On the other hand, CrossEntropy measures the divergence of this prediction from the distribution of target scores. This work introduces the ISBE functionality, justifying the thesis about the redundancy of cross-entropy computation in deep learning of classifiers. Not only can we omit the calculation of entropy, but also, during back-propagation, there is no need to direct the error to the normalization unit for its backward transformation. Instead, the error is sent directly to the model's network. Using examples of perceptron and convolutional networks as classifiers of images from the MNIST collection, it is observed for ISBE that results are not degraded with SoftMax only but also with other activation functions such as Sigmoid, Tanh, or their hard variants HardSigmoid and HardTanh. Moreover, savings in the total number of operations were observed within the forward and backward stages. The article is addressed to all deep learning enthusiasts but primarily to programmers and students interested in the design of deep models. For example, it illustrates in code snippets possible ways to implement ISBE functionality but also formally proves that the SoftMax trick only applies to the class of dilated SoftMax functions with relocations. [ABSTRACT FROM AUTHOR]

Copyright of Ensayos - Revista de Economía is the property of Ensayos Revista de Economia and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Refrigeration is the property of Journal of Refrigeration Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The study presents the utilization of an innovative neutrosophic cross entropy-based technique for the discrimination of the antioxidant potential of Arisaema tortuosum leaf extract based on polyphenolic content. The effect of three extraction techniques, namely, Soxhlet extraction, ultrasound-assisted extraction, and maceration, have been analyzed on the percentage yield of the extract. The effect of solvents, namely, methanol, chloroform and hexane, on the percentage yield has also been explored. The leaf extract was found to exhibit significant antioxidant potential and polyphenolic content. Substantial discrimination was observed among antioxidant potential and polyphenolic contents with minimum neutrosophic cross entropy values designated to the extant higher amount of polyphenols in A. tortuosum leaf extract. The proposed neutrosophic cross entropy-based technique is beneficial for further mathematical treatments because of its symmetric nature in comparison with the existing methods, which may indicate vagueness in the evaluation information under certain situations and thus affects the prognosis analysis.

Aiming at the problem of "fast and precise" in the optimization of joint fire strike scheme, a cross-entropy-quantum particle swarm optimization algorithm is designed, and an intelligent optimization method of joint fire strike scheme based on this algorithm is proposed. The algorithm fully combines the good global optimization ability of the cross-entropy algorithm and the efficient operation of the quantum particle swarm algorithm. By constructing a discrete probability matrix, the uncertain motion of the particles is more directional and the cooperative optimization ability of the particles is improved. The experimental results show that the intelligent optimization method of the joint fire strike scheme based on this algorithm can effectively improve the optimization effect and optimization efficiency of the joint fire strike scheme, and has obvious advantages compared with the standard cross-entropy algorithm and the standard quantum particle swarm optimization algorithm.

The study presents the utilization of an innovative neutrosophic cross entropy-based technique for the discrimination of the antioxidant potential of Arisaema tortuosum leaf extract based on polyphenolic content. The effect of three extraction techniques, namely, Soxhlet extraction, ultrasound-assisted extraction, and maceration, have been analyzed on the percentage yield of the extract. The effect of solvents, namely, methanol, chloroform and hexane, on the percentage yield has also been explored. The leaf extract was found to exhibit significant antioxidant potential and polyphenolic content. Substantial discrimination was observed among antioxidant potential and polyphenolic contents with minimum neutrosophic cross entropy values designated to the extant higher amount of polyphenols in A. tortuosum leaf extract. The proposed neutrosophic cross entropy-based technique is beneficial for further mathematical treatments because of its symmetric nature in comparison with the existing methods, which may indicate vagueness in the evaluation information under certain situations and thus affects the prognosis analysis. [ABSTRACT FROM AUTHOR]

In multi-label learning, each instance simultaneously associates with multiple labels, which means that labeling such instances is quite costly. Active learning, as an important machine learning paradigm, learns the classification model by querying merely a small portion of data with important information, by means of which, the labeling cost can be greatly reduced during the training process and an accurate and robust classification model could be obtained. Therefore, multi-label active learning (MLAL) has garnered increasing attentions. The primary challenge in MLAL lies in designing an effective query strategy to measure uniform information about unlabeled instances throughout all labels. In this study, we propose a query strategy named predicted label vectors inconsistency and cross entropy measure (PLVI-CE) that considers both uncertainty and diversity measures. In PLVI-CE, the uncertainty is measured by the inconsistency between two predicted label vectors from the same unlabeled instance, and the diversity is assessed by the average discrepancy in posterior probabilities between each unlabeled instance and all instances in the labeled set. Furthermore, in this study, we try to adopt label-weighted extreme learning machine (LW-ELM) as the base classifier in the MLAL framework with considering its following advantages: (1) LW-ELM has a low computational cost, (2) LW-ELM has strong generalization performance, and (3) LW-ELM can be directly used to classify multi-label data with class imbalance distributions, hence providing approximately unbiased instance querying during MLAL. Experimental results on 12 benchmark multi-label datasets indicate the effectiveness and superiority of the proposed PLVI-CE algorithm in comparison with several current state-of-the-art MLAL algorithms. [ABSTRACT FROM AUTHOR]

Aiming at the problem of "fast and precise" in the optimization of joint fire strike scheme， a cross-entropy-quantum particle swarm optimization algorithm is designed， and an intelligent optimization method of joint fire strike scheme based on this algorithm is proposed. The algorithm fully combines the good global optimization ability of the cross-entropy algorithm and the efficient operation of the quantum particle swarm algorithm. By constructing a discrete probability matrix， the uncertain motion of the particles is more directional and the cooperative optimization ability of the particles is improved. The experimental results show that the intelligent optimization method of the joint fire strike scheme based on this algorithm can effectively improve the optimization effect and optimization efficiency of the joint fire strike scheme， and has obvious advantages compared with the standard cross-entropy algorithm and the standard quantum particle swarm optimization algorithm. [ABSTRACT FROM AUTHOR]

Terahertz time-domain spectroscopy (THz-TDS) has been widely used for food and drug identification. The classification information of a THz spectrum usually does not exist in the whole spectral band but exists only in one or several small intervals. Therefore, feature selection is indispensable in THz-based substance identification. However, most THz-based identification methods empirically intercept the low-frequency band of the THz absorption coefficients for analysis. In order to adaptively find out important intervals of the THz spectra, an interval-based sparse ensemble multi-class classifier (ISEMCC) for THz spectral data classification is proposed. In ISEMCC, the THz spectra are first divided into several small intervals through window sliding. Then the data of training samples in each interval are extracted to train some base classifiers. Finally, a final robust classifier is obtained through a nonnegative sparse combination of these trained base classifiers. With l1 -norm, two objective functions that based on Mean Square Error (MSE) and Cross Entropy (CE) are established. For these two objective functions, two iterative algorithms based on the Alternating Direction Method of Multipliers (ADMM) and Gradient Descent (GD) are built respectively. ISEMCC transforms the problem of interval feature selection and decision-level fusion into a nonnegative sparse optimization problem. The sparse constraint ensures only a few important spectral segments are selected. In order to verify the performance of the proposed algorithm, comparative experiments on identifying the origin of Bupleurum and the harvesting year of Tangerine peel are carried out. The base classifiers used by ISEMCC are Support Vector Machine (SVM) and Decision Tree (DT). The experimental results demonstrate that the proposed algorithm outperforms six typical classifiers, including Random Forest (RF), AdaBoost, RUSBoost, ExtraTree, and the two base classifiers, in terms of classification accuracy.

This paper proposes a course knowledge resource extraction method based on the improved Hidden Markov Model and combines the method with hierarchical classification ideas. The improved Hidden Markov Model is used in the study of the text classification process of course knowledge resources, the expected cross entropy selects the feature words, and the semantic space composition is obtained by using the implicit semantic indexing method. When setting the observation output probability matrix for the HMM classifier model, the improved TFIDF method is introduced to reflect the semantic relations between feature words. An evaluation criterion such as macro-micro mean is used to analyze the performance of the whole text set classification and gradually improve the HMM model. Finally, the HMM resource information extraction model has been improved to mine the elements of civics in digital electronics technology courses. It is found that in the compulsory textbook knowledge of digital electronics technology, the trigger knowledge has the greatest integration degree with the Civic-Political elements with a percentage of 0.3382, followed by the analog-digital and digital-analog converter knowledge with a percentage of 0.2818. In the textbook column on digital electronics technology, the extended knowledge has the greatest integration degree with the Civic-Political elements with a percentage of 0.5043, followed by the strengthened category column with a percentage of 0.1712. The analysis of the students’ different Civic-Political dimensions after the instruction has been done. In the political identity category, national sentiment, scientific spirit and ecological environment topics improved by 7.23, 2.01, 3.2 and 2.14 points, respectively. It shows that students can improve their cultural knowledge in digital electronics and establish a correct worldview, life view, and values.

The Controller Placement Problem (CPP) is a key research topic in Software Defined Network (SDN), as the communication delay is influenced by the position of controllers and switches. On that basis, the network failures may happen occasionally, which can cause the increase of propagation latency and the reduction of network performance. As a result, it is essential to research the Controller Placement problem for Link Failures (CPLF). In this paper, authors propose a method based on the cross entropy to solve CPP after link failures, and adopt the Halton sequence to reduce the computation overhead of simulating link failures while guaranteeing the accuracy. In the experiments, we measure and compare the worst-case delay among three methods: our proposed cross entropy-based controller placement algorithm, the optimized controller placement algorithm and a greedy-based controller placement algorithm, and conduct experiments on six real network topologies. The experimental results verify that our proposed method can reduce the worst-case delay by 2 7 % in comparison with GPA. Moreover, the proposed method can always find optimized controller placement schemes no matter how the network scale or the number of controller varies, with a less than 0. 6 % error when compared with the optimal solution. [ABSTRACT FROM AUTHOR]

Introduction: The current method of monitoring sleep disorders is complex, time-consuming, and uncomfortable, although it can provide scientifc guidance to ensure worldwide sleep quality. This study aims to seek a comfortable and convenient method for identifying sleep apnea syndrome. Methods: In this work, a one-dimensional convolutional neural network model was established. To classify this condition, the model was trained with the photoplethysmographic (PPG) signals of 20 healthy people and 39 sleep apnea syndrome (SAS) patients, and the influence of noise on the model was tested by anti-interference experiments. Results and Discussion: The results showed that the accuracy of the model for SAS classifcation exceeds 90%, and it has some antiinterference ability. This paper provides a SAS detection method based on PPG signals, which is helpful for portable wearable detection. [ABSTRACT FROM AUTHOR]

An improved cross-entropy (CE) method assisted with a chaotic operator (CGSCE) is presented for solving the optimal power flow (OPF) problem. The introduction of the chaotic operator helps to enhance the exploration capability of the popular cross-entropy approach while the global best solution is preserved. To handle the constraints in the optimal power flow, an efficient constraint handling technique with no parameter adjustment is also introduced. The approach is tested on both the IEEE-30 bus system and the IEEE-57 bus system with different objective functions to verify its effectiveness in comparison with a few other methods reported in the literature. Simulation results confirm that the proposed method is capable of improving both the exploration ability and the convergence speed of the conventional cross-entropy method. It outperforms the original cross-entropy, its variant GSCE and other methods in most of the OPF study cases. [ABSTRACT FROM AUTHOR]

The properties of cardio-respiratory coupling (CRC) are affected by various pathological conditions related to the cardiovascular and/or respiratory systems. In heart failure, one of the most common cardiac pathological conditions, the degree of CRC changes primarily depend on the type of heart-rhythm alterations. In this work, we investigated CRC in heart-failure patients, applying measures from information theory, i.e., Granger Causality (GC), Transfer Entropy (TE) and Cross Entropy (CE), to quantify the directed coupling and causality between cardiac (RR interval) and respiratory (Resp) time series. Patients were divided into three groups depending on their heart rhythm (sinus rhythm and presence of low/high number of ventricular extrasystoles) and were studied also after cardiac resynchronization therapy (CRT), distinguishing responders and non-responders to the therapy. The information-theoretic analysis of bidirectional cardio-respiratory interactions in HF patients revealed the strong effect of nonlinear components in the RR (high number of ventricular extrasystoles) and in the Resp time series (respiratory sinus arrhythmia) as well as in their causal interactions. We showed that GC as a linear model measure is not sensitive to both nonlinear components and only model free measures as TE and CE may quantify them. CRT responders mainly exhibit unchanged asymmetry in the TE values, with statistically significant dominance of the information flow from Resp to RR over the opposite flow from RR to Resp, before and after CRT. In non-responders this asymmetry was statistically significant only after CRT. Our results indicate that the success of CRT is related to corresponding information transfer between the cardiac and respiratory signal quantified at baseline measurements, which could contribute to a better selection of patients for this type of therapy. [ABSTRACT FROM AUTHOR]

Well-Being analysis is an approach that integrates deterministic criteria with probabilistic methods, and it plays a crucial role in the operational planning of power systems. However, assessing the Well-Being of composite generation and transmission systems presents a formidable challenge, characterized by significant computational burdens and sluggish processing speeds. To tackle this issue, we embarked on an effort to enhance the computational efficiency of Well-Being assessment by employing the cross-entropy method (CEM). Nonetheless, our experimental pursuits revealed that the conventional employment of CEM for Well-Being assessment can lead to protracted convergence of the marginal index. To overcome this limitation, we introduce an enhanced multi-objective cross-entropy method (MCEM) that integrates weight factors, thereby ensuring an accelerated convergence rate for both the risk and marginal indices. To validate the effectiveness and advancement of our proposed MCEM approach, we conduct a comprehensive comparative analysis using the IEEE RTS79 and MRTS79 test systems as case studies. We contrast our method with the conventional MCS and CEM approaches, conducting a thorough examination of the computational performance of MCEM. This comprehensive comparative study unequivocally confirms the efficacy and progressive nature of the MCEM framework presented in this paper.

Aiming at the low accuracy and efficiency of stochastic model updating, a stochastic finite element model updating method based on the Kriging model and wavelet packet energy spectrum was proposed. Firstly, assume that the parameters and response characteristics of the model to be updated obey normal distributions, the uncertainty model updating was transformed into the updating of mean and standard deviation. Secondly, the parameters to be updated were taken as inputs of Kriging model, the node energies extracted by the acceleration frequency response function after wavelet packet decomposition were taken as the outputs, the political optimizer algorithm was introduced to optimize the correlation coefficient to construct Kriging model. Then, minimize the absolute value of the difference between the test response and the predicted response as the objective function for updating mean, and minimize the cross entropy as the objective function for updating standard deviation, and updated the parameters mean and standard deviation through the political optimizer algorithm. Finally, taking a space truss structure as the example, the elastic modulus and density were selected as the parameters to be updated to verify the feasibility of the proposed method. The results show that the proposed method can effectively update the mean and standard deviation of structural parameters, and errors of the updated mean and standard deviation are less than 0.1% and 3.5%, respectively.

In order to effectively evaluate the grinding performance of CBN（Cubic Boron Nitride） grinding wheel in the grinding process, a monitoring method of CBN grinding wheel performance was proposed based on Shannon entropy theory and acoustic emission signal. Firstly, the acoustic emission signal of CBN grinding wheel was collected by an acoustic emission sensor. Then, the probability density and distribution of acoustic emission signal of CBN grinding wheel was estimated by using the maximum information entropy. In addition, the characteristics of acoustic emission signals of CBN grinding wheel during cyclic grinding after dressing and residual grinding with different diameters were analyzed and discussed. Subsequently, the relationship between the maximum entropy probability density distribution of acoustic emission signals and the different grinding performance of CBN grinding wheel was established according to the cross entropy principle. Finally, a large number of experiments verified the availability above mentioned estimation method. The CBN grinding wheel wear status and the remaining life of CBN grinding wheel can be monitored.

Medical imaging description and disease diagnosis are vitally important yet time-consuming. Automated diagnosis report generation (DRG) from medical imaging description can reduce clinicians’ workload and improve their routine efficiency. To address this natural language generation task, fine-tuning a pre-trained large language model (LLM) is cost-effective and indispensable, and its success has been witnessed in many downstream applications. However, semantic inconsistency of sentence embeddings has been massively observed from undesirable repetitions or unnaturalness in text generation. To address the underlying issue of anisotropic distribution of token representation, in this study, a contrastive learning penalized cross-entropy (CLpCE) objective function is implemented to enhance the semantic consistency and accuracy of token representation by guiding the fine-tuning procedure towards a specific task. Furthermore, to improve the diversity of token generation in text summarization and to prevent sampling from unreliable tail of token distributions, a diversity contrastive search (DCS) decoding method is designed for restricting the report generation derived from a probable candidate set with maintained semantic coherence. Furthermore, a novel metric named the maximum of token repetition ratio (maxTRR) is proposed to estimate the token diversity and to help determine the candidate output. Based on the LLM of a generative pre-trained Transformer 2 (GPT-2) of Chinese version, the proposed CLpCE with DCS (CLpCEwDCS) decoding framework is validated on 30,000 desensitized text samples from the “Medical Imaging Diagnosis Report Generation” track of 2023 Global Artificial Intelligence Technology Innovation Competition. Using four kinds of metrics evaluated from n-gram word matching, semantic relevance, and content similarity as well as the maxTRR metric extensive experiments reveal that the proposed framework effectively maintains semantic coherence and accuracy (BLEU-1, 0.4937; BLEU-2, 0.4107; BLEU-3, 0.3461; BLEU-4, 0.2933; METEOR, 0.2612; ROUGE, 0.5182; CIDER, 1.4339) and improves text generation diversity and naturalness (maxTRR, 0.12). The phenomenon of dull or repetitive text generation is common when fine-tuning pre-trained LLMs for natural language processing applications. This study might shed some light on relieving this issue by developing comprehensive strategies to enhance semantic coherence, accuracy and diversity of sentence embeddings.

In deep learning of classifiers, the cost function usually takes the form of a combination of SoftMax and CrossEntropy functions. The SoftMax unit transforms the scores predicted by the model network into assessments of the degree (probabilities) of an object’s membership to a given class. On the other hand, CrossEntropy measures the divergence of this prediction from the distribution of target scores. This work introduces the ISBE functionality, justifying the thesis about the redundancy of cross-entropy computation in deep learning of classifiers. Not only can we omit the calculation of entropy, but also, during back-propagation, there is no need to direct the error to the normalization unit for its backward transformation. Instead, the error is sent directly to the model’s network. Using examples of perceptron and convolutional networks as classifiers of images from the MNIST collection, it is observed for ISBE that results are not degraded with SoftMax only but also with other activation functions such as Sigmoid, Tanh, or their hard variants HardSigmoid and HardTanh. Moreover, savings in the total number of operations were observed within the forward and backward stages. The article is addressed to all deep learning enthusiasts but primarily to programmers and students interested in the design of deep models. For example, it illustrates in code snippets possible ways to implement ISBE functionality but also formally proves that the SoftMax trick only applies to the class of dilated SoftMax functions with relocations.

Supply chain finance has solved the problem of financing difficulties for small and medium-sized enterprises in the upstream and downstream of China's supply chain. However, with the development of the economy, traditional supply chain finance is gradually unable to meet the financing needs of enterprises. Enterprise financing pursues simplified processing time and process, simple operation methods and procedures. At the same time, the rapid development of information technology and the emergence and prosperity of new technologies such as e-commerce, big data, cloud computing also continue to promote the breeding of new models. Based on this background, a new type of supply chain finance has emerged – Internet supply chain finance. The risk assessment of internet supply chain finance is a classical multiple-attributed decision making (MADM) problems. In this paper, the cross-entropy method under type-2 neutrosophic numbers (T2NNs) is built based on the traditional cross-entropy method. Firstly, the T2NN is introduced. Then, combine the traditional fuzzy cross-entropy method with T2NNs information, the type-2 neutrosophic number cross-entropy (T2NN-CE) method is established for MADM under T2NNs. Finally, a numerical example for risk assessment of internet supply chain finance has been given and some comparisons is used to illustrate advantages of T2NN-CE method with T2NNs. [ABSTRACT FROM AUTHOR]

Deep learning technology is widely used in the field of number recognition.It constructs neural network model through deep learning technology,nonlinear transformation activation function in neurons,different activation functions with different parameter initialization strategies,trains MINIST handwritten data set,constructs analysis model and recognizes numbers in images,reduce the dimension of a large amount of data into a small amount of data,and ensure the effective retention of image features.Through the analysis of image data,adding the feature conversion process,using the gradient descent optimizer to build a network structure and reduce the dimension of data,which can effectively avoid over fitting.Cross-entropy verification is used to compile and train the model,and the output classification results are further analyzed.Through the K-nearest neighbor classification algorithm,KNN classifier is set to further improve the accuracy of classification and prediction.Through MNIST data set experiment,the recognition rate is about 96.2%.The K-nearest neighbor algorithm(KNN) is introduced into the output layer,combined with the full connection layer and softmax layer of traditional convolutional neural network(CNN).After cross verification,the recognition rate is 99.6%.

Tri-training expands the training set by adding pseudo-labels to unlabeled data, which effectively improves the generalization ability of the classifier, but it is easy to mislabel unlabeled data into training noise, which damages the learning efficiency of the classifier, and the explicit decision mechanism tends to make the training noise degrade the accuracy of the classification model in the prediction stage. This study proposes the Tri-training algorithm for adaptive nearest neighbor density editing and cross-entropy evaluation (TTADEC), which is used to reduce the training noise formed during the classifier iteration and to solve the problem of inaccurate prediction by explicit decision mechanism. First, the TTADEC algorithm uses the nearest neighbor editing to label high-confidence samples. Then, combined with the relative nearest neighbor to define the local density of samples to screen the pre-training samples, and then dynamically expand the training set by adaptive technique. Finally, the decision process uses cross-entropy to evaluate the completed base classifier of training and assign appropriate weights to it to construct a decision function. The effectiveness of the TTADEC algorithm is verified on the UCI dataset, and the experimental results show that compared with the standard Tri-training algorithm and its improvement algorithm, the TTADEC algorithm has better classification performance and can effectively deal with the semi-supervised classification problem where the training set is insufficient. [ABSTRACT FROM AUTHOR]

We introduce a cross-entropy (CE) indicator to quantify the extent to which two input–output tables or two tables with results based on input–output analysis differ from each other. Our work deploys a unique feature of the CE indicator: it can be decomposed, allowing for matrix comparisons at various levels within one coherent framework. To illustrate the power of this approach, we apply the technique to five multi-region input–output (MRIO) tables for 2011, derived from the Eora, EXIOBASE, GTAP, OECD and WIOD databases. We make pairwise comparisons between MRIOs and between global value chain (GVC) computations based on these MRIOs. We find that answers to questions related to broader aggregates are generally quite similar, but that answers to questions at the level of single industries can be rather different across MRIOs. [ABSTRACT FROM AUTHOR]

To keep the main system and the auxiliary equipment of production system of thermal power plants operating normally, it's a necessary work to fully consider the periodical repeat features of thermal process when modelling the thermal process based on thermal state data of machine units. Since most equipment in power plant work in an electromagnetic coupling environment, a too high ambient temperature will affect the safe operation of units, so when studying the dynamic features of multi-modal thermal process of power plants and their soft environment, the temperature distribution features of multimodal thermal process should be classified. To solve these questions, this paper established a model for the coupled temperature field during multi-modal thermal process and studied its analytical solutions. At first, this paper proposed a multi-modal modelling method for solving the distribution features of coupled temperature during multi-modal thermal process, the method can quickly determine the optimal number of local models of multimodal thermal process of power plant units while reducing the computational load of the coupled temperature field model. Then, this paper introduced an improved niche method into the basic cross-entropy optimization algorithm and used it to solve the layout optimization problem of coupled temperature field during multi-modal thermal process. At last, the validity of the proposed method was verified by experimental results. [ABSTRACT FROM AUTHOR]

Medical pictures play a vital role in novel diagnosis; to achieve this, the quality of these images must be improved in a way that meets the necessary standards. To deal with this matter, we have proposed an optimal new histogram equalization technique with less computational complexity. In this way, using an automated thresholding technique, the image-histogram is isolated into two sub-histograms, and to avoid undesirable effects of conventional methods which cause abnormal brightness and intensity saturation effects, we proposed a new composition of Contrast-limited adaptive histogram equalization and Brightness preserving dynamic fuzzy histogram equalization techniques with appropriate weights according to the histogram situation based on a fuzzy approach. To improve the image contrast, pixels whose luminous intensity is higher than the proposed calculated floating threshold are equalized. Then, they are combined with the lower bound of the histogram and finally the normalization is applied on them. In order to evaluate the consequences, 5 different authenticated medical image databases containing mammography, ultrasonic, Computed Tomography scan and Magnetic Resonance Imaging pictures are used. Using effective quality measurements, e.g., Peak Signal to Noise Ratio, Mean Square Error, Absolute Means Brightness Error, the Index of Structural Similarity and Measure of Enhancement, significant outcomes have been obtained in different databases. The proposed algorithms are advantageous in brightness preserving and contrast enhancement of different medical images. These methods improve the quality of medical pictures and help doctors in the diagnosis process. [ABSTRACT FROM AUTHOR]

This study introduces a novel fault recognition methodology for turbine faults through symmetric trigonometric fuzzy and neutrosophic cross entropy measures (FCEM and NCEM) consequently. After knowing the nethermost (lowest) and uppermost (highest) energy bounds of each real fault conditions, the energy interval ranges are constructed and then transformed into the form of single valued neutrosophic (SN) sets. Thereafter, the proposed symmetric trigonometric cross -entropy measures are deployed to recognize faults of turbine. The nethermost FCEM and NCEM values between familiar and unfamiliar fault conditions indicates that the unfamiliar fault condition is closer to the familiar one. The applicability of the proposed methodology is validated by taking into consideration the example of fault diagnosis of turbine. The repercussions of this study yield that the proposed symmetric trigonometric FCEM and NCEM cannot only recognize optimal fault, they can also provide meaningful and remarkable fault information. A comparison of the underlying FCEM and NCEM (based on SN sets) with the enduring cosine measures (based on vague sets) conclude that the latter sets may hide some fruitful fault information., when experimented under sensitive and intuitive criteria and thus resulting an incomplete fault evaluation criterion

The underlying study intends to evaluate community health risk assessment from fluoride contamination of groundwater samples employing the proposed entropy variants of single valued neutrosophic sets. The symmetric fuzzy cross entropy numbers, which can represent the macroscopic view of fluoride contamination more effectively, are constructed in this study and then deployed to rank the seasonal parameters (pre-monsoon, rainy season and pre-monsoon) responsible for fluoride contamination in the study area. To quantify the non-linear relationship between seasonal parameters and sampling spots, the proposed neutrosophic entropy variants are fascinated for assigning weights to the monitored concentration reading of each seasonal parameter with respect to various sampling spots. Thereafter, these weights are coupled with the quality rating scale of each parameter, intended to establish new fuzzy and single valued neutrosophic entropy weighted fluoride contamination indices (FEFCI & NEFCI) respectively. The maximum (or minimum) FEFCI or NEFCI score at a particular sampling spot is designated to the “most (or least) contaminated” sampling spot accordingly.The underlying fluoride contaminated sampling spot identification methodology is efficacious for providing a better insight in assessing the community health risk from ground water of the study area.

To address the problem of low detection accuracy due to training noise caused by mislabeling when Tri-training for network intrusion detection (NID), we propose a Tri-training algorithm based on cross entropy and K-nearest neighbors (TCK) for network intrusion detection. The proposed algorithm uses cross-entropy to replace the classification error rate to better identify the difference between the practical and predicted distributions of the model and reduce the prediction bias of mislabeled data to unlabeled data; K-nearest neighbors are used to remove the mislabeled data and reduce the number of mislabeled data. In order to verify the effectiveness of the algorithm proposed in this paper, experiments were conducted on 12 UCI datasets and NSL-KDD network intrusion datasets, and four indexes including accuracy, recall, F-measure and precision were used for comparison. The experimental results revealed that the TCK has superior performance than the conventional Tri-training algorithms and the Tritraining algorithms using only cross-entropy or K-nearest neighbor strategy. [ABSTRACT FROM AUTHOR]

The task of Point-of-Interest (POI) recommendation is to recommend the next interest locations for users. Gated Graph Neural Network (GGNN) has been proved to be effective on POI recommendation tasks. However, existing GGNN solutions rarely consider the spatiotemporal information between nodes in the sequence graph, which is essential for modeling user check-in behaviors in next POI recommendation. In this paper, we propose an attention-based spatiotemporal gated graph neural network model (ATST-GGNN) for next POI recommendation. Firstly, the user’s check-in sequence is represented as a graph structure. Secondly, we use spatiotemporal context information to dynamically update nodes in the sequence graph, and obtain the complex transfer relationships between the check-ins. Thirdly, each session is then represented as the composition of the long and short preference using an attention network. However, current short preference fails to model union-level sequential patterns, we improve the local embedding representation of graph nodes by window pooling method, as well as the global embedding representation of graph nodes by integrating it into attention mechanism. Finally, the objective function is constructed by cross entropy and the model parameters are learned. The experimental results show that the precision rate and mean reciprocal ranking of ATST-GGNN method are greatly improved compared with the state-of-art methods. It has good application prospect.

Recent study shows that recommendation system not only relys on user’s static preference, but also dynamic preference. Consequently, it leads to the emergence of session-based recommendation. With the development of recurrent neural network, this kind of method can capture representations of users’ sequential behaviors from a large number of sessions. However it is prone to spurious dependency problem. Recently, convolutional neural network has also shown its potential in modelling session, especially in extracting complex local pattern of subsequence. Therefore, we propose a hybrid neural model, called SGPD, for learning sequential general pattern and dependency for session-based recommendation. In SGPD, we propose recurrent residual convolution network to extract general pattern of subsequence in a session. Furthermore, the SGPD scans sequence from forward and reverse direction by bidirectional recurrent neural network, and learns sequential dependency of a session. Finally, the objective function is constructed by cross entropy and the model parameters are learned. The experimental results show that the precision rate, recall rate and mean reciprocal ranking of SGPD are greatly improved compared with the state-of-art methods. It has good application prospect.

In the last decades simulation methods for calculating failure probabilities in structural reliability have become increasingly popular. The FORM/SORM concepts were discarded based on the claim that these methods do not work well for high dimensional problems and for multiple design points. From this it was concluded that design points are quite useless. This is based a confusion between the design point concept and the FORM/SORM approximation method. The Lemma of Hohenbichler shows that for small probabilities the probability content of the failure domain is concentrated around the design points, so these points are important. Their positions give further the conditional means of the points in the failure domains. Then it is proven that most of the alternative estimation methods, surrogate models, cross entropy, subset simulation and line sampling use implicitly design point concepts. All these approaches end up with a point clouds around the design points. Finally design point based methods are derived which are efficient for high dimensions and multiple design points. • Structural reliability is analyzed using structuralist philosophy. • Conditional means and variances in the failure domain are asymptotic functions of the design points. • Surrogate modeling, cross entropy estimation, important direction and subset simulation all end up at the design points. • New design point based estimation method is described in section 9 using partitions of unity. [ABSTRACT FROM AUTHOR]

An improved cross-entropy (CE) method assisted with a chaotic operator (CGSCE) is presented for solving the optimal power flow (OPF) problem. The introduction of the chaotic operator helps to enhance the exploration capability of the popular cross-entropy approach while the global best solution is preserved. To handle the constraints in the optimal power flow, an efficient constraint handling technique with no parameter adjustment is also introduced. The approach is tested on both the IEEE-30 bus system and the IEEE-57 bus system with different objective functions to verify its effectiveness in comparison with a few other methods reported in the literature. Simulation results confirm that the proposed method is capable of improving both the exploration ability and the convergence speed of the conventional cross-entropy method. It outperforms the original cross-entropy, its variant GSCE and other methods in most of the OPF study cases.

In order to overcome the spectral interference of the conventional Fourier transform in the International Electrotechnical Commission framework, this paper introduces a Bregman-split-based compressive sensing (BSCS) method to estimate the Taylor–Fourier coefficients in a multi-frequency dynamic phasor model. Considering the DDC component estimation, this paper transforms the phasor problem into a compressive sensing model based on the regularity and sparsity of the dynamic harmonic signal distribution. It then derives an optimized hybrid regularization algorithm with the Bregman split method to reconstruct the dynamic phasor estimation. The accuracy of the model was verified by using the cross entropy to measure the distribution differences of values. Composite tests derived from the dynamic phasor test conditions were then used to verify the potentialities of the BSCS method. Simulation results show that the algorithm can alleviate the impact of dynamic signals on phasor estimation and significantly improve the estimation accuracy, which provides a theoretical basis for P-class phasor measurement units (PMUs). [ABSTRACT FROM AUTHOR]

Copyright of Journal of Zhengzhou University (Natural Science Edition) is the property of Journal of Zhengzhou University (Natural Science Edition) Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)