Search

One of the big challenges of using knowledge graphs for recommendations is how to capture the structured know-ledge of items and extract its semantic features. To solve this problem, this paper proposed a collaborative filtering recommendation algorithm based on knowledge graph embedding（KGECF）. Firstly, it extracted the knowledge information related to items from the Freebase knowledge graph and linked with historical interactive items to construct knowledge subgraphs. Then it obtained the representation of entities and relations in the sub-graph using the Xavier-TransR method based on TransR. This paper designed an end-to-end joint learning model to embed structured information and historical preference information into a unified vector space. Finally, it used the collaborative filtering method to further calculate these vectors to generate an accurate re-commendation list. Experimental results on two public datasets MovieLens-1 M and Amazon-book show that the proposed algorithm is superior to the baseline algorithms in terms of precision, recall, F1 value and NDCG metrics. It means that the above method can integrate large-scale structured and unstructured data, while obtaining high precision recommendation results. [ABSTRACT FROM AUTHOR]

In order to solve the problem that the original sparrow search algorithm (Sparrow Search Algorithm, SSA) has reduced population diversity and the weak local development ability makes it difficult to jump out of the local optimum under the condition of approaching the global optimum. This paper proposes a sparrow search algorithm based on tangential flight (tangent flight Sparrow Search Algorithm, tanSSA). First of all, using adaptive t-distribution strategy to improve the finder location update formula can improve the individual sparrow’s optimization ability and prevent the algorithm from maturing. Then, the tangent flight strategy in the tangent search algorithm can enhance the search space ability of the algorithm, and can make the algorithm jump out of the advantage of the local optimal solution. In original sparrow search algorithm, using the tangent flight disturbance strategy to disturb the optimal solution. The combination of these two strategies can effectively improve the exploration and development performance of tanSSA algorithm. Finally, using 12 standard benchmark functions to test and verify the optimization performance of tanSSA algorithm combined with Wilcoxon rank sum test, and compared with the original SSA algorithm, whale optimization algorithm, particle swarm optimization algorithm and adaptive t-distribution SSA algorithm. The experimental results show that the optimization ability and convergence speed of the sparrow search algorithm based on tangent flight have a significant improvement. [ABSTRACT FROM AUTHOR]

The three-dimensional path planning problem is a combinatorial optimization problem to find the optimal path be- tween the starting point and the destination in the interference environment. Aiming at the shortcomings of traditional swarm in- telligence algorithm in solving this problem,such as low convergence accuracy and easy to fall into local optimum, this paper proposed an adaptive moth-flame optimization algorithm to solve the problem. By introducing the dynamic adjustment strategy of the flight direction and the position intersection strategy,the improved algorithm continuously generated new individuals while dynamically adjusting the flight direction of the moth, which effectively avoided the algorithm from falling into local optimum ; by adaptively adjusting the number of flames, so as to the algorithm enhanced population diversity in the global exploration stage a- void premature convergence. It used the adaptive moth-flame optimization algorithm and other swarm intelligence algorithms to solve the three-dimensional path planning problem. The experimental results show that the improved adaptive moth-flame opti- mization algorithm has the smallest cost value and the fastest convergence speed among all algorithms. The algorithm has better solution ability in 3D path planning problem. [ABSTRACT FROM AUTHOR]

The current ADMET classification methods have shortcomings in classifying the ADMET of compounds data with multiple characteristics and feature correlation. Moreover, the classification results of ADMET are not explainable. To solve these problems, this paper proposed a classification model based on capsule network（CapsMC). CapsMC model first proposed the Feature-to-Image algorithm. It used this algorithm to consider the correlation and dependence between features into the classification basis, and realized the multi-level extraction of features. Second, it explored the advanced application of capsule network, and designed a cognitive reasoning mechanism. It used this mechanism to carry out cognitive reasoning on features, and realized explainable classification of ADMET. Experimental results on five ADMET datasets show that CapsMC can achieve the explainable classification of ADMET well. [ABSTRACT FROM AUTHOR]

There are two main problems in zero-shot object detection—the correspondence between the visual features of the object and its corresponding category semantic information is not accurate, and the discrimination between the unseen category object and the background is confused. The ACDG algorithm proposed adopts the adaptive weighting mechanism of feature map channels, which can automatically assign weights to feature map channels according to significance, to establish more accurate visual-semantic correspondence relations. In order to solve the problem of low discrimination between the unseen objects and the background, the method introduces the Center Distance Gaussian Distribution Loss to constrain the distance between the center point of the boundary frame and the groundtruth, thus speeding up the convergence of the loss function. In order to verify the superiority of the proposed algorithm, a large number of experiments were performed on COCO dataset, and the recall rate and average accuracy were 5.9% and 4.5% higher than the baseline, respectively. The method presented in this paper has positive significance for improving the performance of zero-shot object detection. [ABSTRACT FROM AUTHOR]

Supervised super-resolution algorithms require plentiful paired-image for networks training, resulting in poor visual recovery effects when processing real-world low-resolution images. To solve this problem, this paper proposed an improved semi supervised algorithm based on CycleGAN, donated as Cycle-SRNet. Firstly the degradation model generated images similar to real world low-resolution faces for network training, and then the reconstruction model restored high resolution face image. Furthermore, this paper introduced perceptual loss function to ensure the structural similarity of the faces so as to recover the facial features better. Experimental results show that the algorithm can restore fuzzy low-resolution video surveillance face images into recognizable face images, and it exceeds SRCNN, SRGAN and CinCGAN in terms of FID, PSNR and SSIM indicators. [ABSTRACT FROM AUTHOR]

Disaster preparedness measures can prepare for disaster relief to ensure that emergency supplies can reach the disaster area in a timely and effective manner. This paper proposed a bi-level programming emergency resource scheduling location-routing model with considering preparedness ( BLERSLR-P) In the model, the upper level aimed to minimize the total cost of supply station construction and operation, while the lower level aimed to minimize the cost of distribution route. This paper designed an improved bi-level salp swarm genetic algorithm (BLSSAGA) to solve this problem, combined the concept of iterative division to update the leader position, used the natural exponential inertia weight strategy to modify the control factor, used the chaotic map to update the follower position, and used Taguchi analysis to obtain reasonable values of parameters. Finally, this paper used bi-level salp swarm genetic algorithm ( BLSSAGA) and genetic particle swarm optimization algorithm ( GAPSO), particle swarm optimization algorithm( PSO), and immune optimization algorithm(IA ) to solve and compare and analyze the LRP data set in OR-Library, which verified the feasibility and effectiveness of the proposed model and algorithm. [ABSTRACT FROM AUTHOR]

In order to improve the accuracy and quality of the personalized recommendation system, this paper constructed a collaborative filtering recommendation algorithm based on SVD and intuitionistic fuzzy clustering ( SVDIFC-CF) to address the information overload and data sparsity problems in traditional recommendation algorithms. The algorithm firstly introduced SVD to fill in the original matrix after dimensionality reduction. Then it used the user-item preference matrix to perform intuitionistic fuzzy clustering of users. Finally it calculated the top N users with the highest similarity to the target user, and found the items that users were most interested in as recommendation results. This paper used the Movie Lens and Jester datasets to verify the effectiveness of the algorithm. The experimental results show that compared with the traditional recommendation algorithm, this algorithm can effectively solve the problem of data sparseness and cold start, and improve the accuracy and quality of recommendation. [ABSTRACT FROM AUTHOR]

In order to reduce the number and total distance of changing parking spaces in a parking lot or a given parking area through autonomous valet parking (AVP), in doing so to lower the related cost and the potential accident risk, this paper formula ted a vehicle-slot matching optimization model which satisfied the shared parking requirement. In view of the feature of autonomous vehicle freely translocating during parking, this paper defined the match, match slice and match map corresponding to decision variable and feasible solution by subdividing the shared parking demand and supply in time. By conception replacement, the paper changed the searching effective neighbor of match map into the classic assignment problem and solved it with Hungarian algorithm. In view of the NP-hard feature of the matching model, this paper proposed a designed tabu searching algorithm. Numerical analysis not only verified the rationality of this model and the effectiveness of the algorithm, also proved that the model and algorithm could deal with the shared parking of conventional vehicles. The results show that AVP can improve the utility of shared parking berths and increase the number of served vehicles. [ABSTRACT FROM AUTHOR]

In order to solve the problems of BBO such as insufficient search scope in the early stage and easy to fall into local optimization in the later stage, this paper proposed an improved El-BBO which based on El. Firstly, the algorithm searched for new habitat around the original habitat, it enhanced the diversity of the initialization population. Secondly, the algorithm made improved habitat expansion, it improved the convergence efficiency of the algorithm. Finally, the algorithm used gradient de scent to make quadratic convergence in the field of optimal solution, which improved the convergence accuracy of the algorithm. This paper carried out 50 Monte Carlo experiments on 12 optimized test functions commonly used in CEC2014, the experimental results show that the overall performance of El-BBO is better than the other three intelligent optimization algorithms in terms of optimal fitness value, average fitness value and standard deviation. It shows that El-BBO can improve the ability to find the optimal solution and enhance the search stability. [ABSTRACT FROM AUTHOR]

In the industrial Internet environment, concerning about poor cooperation efficiency and high production cost of intercell scheduling, this paper solved the problem in the case of overlapping placement of machines. Firstly, the paper used the hierarchical network design idea to construct a two-layer directed processing network with machines and manufacturing cells as nodes. And it proposed a multi-objective scheduling model aiming at minimizing the average degree, completion time and processing cost by analyzing the relationship among the global cooperation efficiency, redundant processing paths and the first-or der degree. Secondly, according to the strong local search ability of sparrow search algorithm, this paper designed a fusion strategy that based on nondominated sorting genetic algorithm and sparrow search algorithm, and an initial solution generation mechanism basing on clustering coefficient. Finally, the experiment illustrates the correlation between network characteristics and intercell scheduling objectives, so the proposed model and algorithm have higher quality of solution. [ABSTRACT FROM AUTHOR]

As collaborative filtering recommendation algorithm relies on user data, there is a great risk of privacy leakage. Differential privacy protection technology can provide strict privacy protection effects, but most recommendation algorithms based on differential privacy do not consider implicit feedback data. To solve the problem, this paper proposed a new collaborative filtering recommendation algorithm based on differential privacy protection. Firstly, it performed matrix decomposition on the implicit feedback matrix to obtain the implicit eigenvectors of users and items. Then it merged the implicit eigenvectors into the solution of explicit feedback model. The algorithm satisfied e-differential privacy protection by adding mean perturbation and gradient perturbation to the model. Finally this paper applied this algorithm to predict the score, and evaluated the effectiveness of the algorithm on the MovieLens dataset. Experimental results show that the algorithm can achieve an effective balance between the accuracy of recommendation results and privacy protection of users. [ABSTRACT FROM AUTHOR]

To solve the problem that the influencing factors of the airport bus operation process are complicated and the operation time is difficult to predict, this paper established an airport bus operation time prediction model based on the subspace identification algorithm. Firstly, based on the multi-source big data generated during the operation process, considering the number of passengers, departure interval, road congestion and other factors at different times, it established the state space model of the airport bus operation. Then it extracted the characteristic variables suitable for describing the operation process of the airport bus as the input and output of the model,and solved the model by subspace identification method. Finally,this paper took an actual operating route of the capital airport bus as a case for simulation analysis. The calculation results show that the mean absolute percentage error and mean square error of the model are 2. 25% and 4. 77, respectively, which are better than the traditional BP neural network prediction model and the least square identification model. The model has good prediction accuracy and has certain practical application value. [ABSTRACT FROM AUTHOR]

Aiming at the problems of low consensus efficiency and insufficient incentive mechanism existing in current consensus algorithms, this paper proposed an alliance chain consensus algorithm based on credit. Firstly, according to the behavior of nodes participating in the consensus process, it designed a node credit evaluation mechanism, and solved the problem of insufficient incentive mechanism between nodes through credit rewards. Secondly, it constructed a credit blockchain and credit calculation model, stored the credit value of the node, and used it as the basis for selecting "miner" nodes, which improved the efficiency of the consensus algorithm. Finally, this paper proposed a round-by-round miner node selection algorithm, which used random algorithm and priority ranking algorithm to select miner nodes in turn, and proposed a node credit evaluation method to avoid excessive credit value of nodes and became "oligarchs" and ensured that the fairness of the nodes becoming miners. The experimental simulation results show that the credit consensus algorithm has low computational power consumption, fast block production speed, better performance than the existing consensus algorithm, and can be well applied to the business and medical alliance chain scenarios. [ABSTRACT FROM AUTHOR]

In order to solve the problems of block effect and boundary artifact in multi focus image fusion, this paper proposed a fusion algorithm combining iterative guided filtering and dictionary learning super-pixel clustering. First, this algorithm divided the source image into super-pixels, which got the clustering of super-pixels using the peak density, and then extracted the feature matrix using the super-pixel cluster blocks as processing units. Then, it established a low rank representation model and introduced Laplacian regular terms to increase the spatial consistency constraints of the same kind neighboring regions. It constructed a low rank representation dictionary and solved the model coefficients by linear alternating iteration with its own quadratic penalty term. This algorithm constructed a preliminary focused feature map by combining the low-rank representation coefficient matrix and the error matrix, and used iterative guided filtering to optimize the focus feature map, it expanded multifocal differentiation. Finally, this paper got the final fusion result based on optimized focusing feature map and source image. Simulation results show that the fusion results of this algorithm are superior to other similar algorithms in terms of both subjective vision and objective indicators. [ABSTRACT FROM AUTHOR]

This paper presented an image dehazing method based on deep learning to solve the problem that the remote sensing images had reduced image sharpness due to haze. Firstly, this algorithm distorted the original atmospheric scattering model to get an end-to-end defrosting model. Then, it unified several unknown parameters to one parameter and estimated the unknown parameters by using multiscale convolution neural network. Finally, it brought the parameter estimates into the defrosting model to get a haze-free image. For no reference image dataset, this paper used the existing dataset to preliminary train the network, then added the self-built dataset to secondary train the network. The experimental results show that, compared with the related defrosting algorithms, the proposed algorithm improves the visual effect and objective index to some extent, and effectively improves the clarity of remote sensing images in haze weather conditions. [ABSTRACT FROM AUTHOR]

In order to solve the problem of high management cost of centralized identity authentication in the application scenario of Internet of Things ( IoT ), and solve the problem of mutual trust and certificate management between heterogeneous trust domains, this paper proposed a c ross- domain authentication scheme based on multi-layer blockchain. This scheme used local blockchain for distributed node management in IoT application scenarios, and used public blockchain for cross chain identity authentication between blockchains. This paper designed a protocol for cross domain authentication. In order to solve the problem of mutual trust among multiple management domains in cross-domain access, this paper proposed a DOPS consensus based on trust evaluation to evaluate the credibility of nodes. Finally, it analysed the security and efficiency of c ross domain authentication scheme. The analysis shows that the scheme has good security and effectiveness, and improves the efficiency of cross domain authentication. [ABSTRACT FROM AUTHOR]

In order to solve the trust problem and resource scheduling efficiency problem in the current cloud manufacturing system, this paper researched the application of block chain technology in cloud manufacturing system. Firstly, this paper expounded the significance of applying blockchain technology to cloud manufacturing system, and proposed a cloud manufacturing system based on blockchain technology. Secondly, this paper designed the manufacturing resource scheduling mode based on smart contract. And constructed the resource scheduling model aim to with the minimum manufacturing cost, the shortest time and the highest qualification rate. The model used differential evolution algorithm to solve. Finally, it simulated the experiment . The simulation results show the resource scheduling method based on the blockchain technology in the smart contract not only ensures the mutual trust between the participants in the system, but also effectively improves the resource scheduling efficiency and the superiority of the resource scheduling scheme of the cloud manufacturing system. [ABSTRACT FROM AUTHOR]

In order to resist the attack of quantum algorithm and deal with the defect that malicious signers can output multiple signatures using the complete anonymity of ring signature technology to carry out double overhead attack, and to solve the unnecessary waste of system overhead, this paper proposed a new identity-based linkable ring signature scheme form lattice. The scheme takes the approximate shortest vector problem on the lattice as the security basis, reduces the solution of the problem to the solution of the collision problem, generates the signature by using the linear operation between matrix vectors, and combines the identity-based cryptography technology. It solved the problem of system overhead waste, does not involve complex algorithms such as trapdoor generation and Gaussian sampling, improves the signature efficiency and reduces the storage overhead. It is verified that the scheme meets the requirements of complete anonymity and strong unforgeability under the random oracle model. After analysis, the scheme is a secure and efficient ring signature scheme. [ABSTRACT FROM AUTHOR]

Most of the existing smart grid authentication schemes suffer from high computational costs and complex authentication processes, which are not suitable for resource-constrained smart devices in the smart grid, while some lightweight schemes suffer from various security vulnerabilities, none of which can achieve the required trade-off between efficiency and security. To solve these problems, this paper designed an enhanced provably secure smart grid lightweight anonymous authentication scheme based on elliptic curve encryption algorithm. Introducing auxiliary validator to get rid of the reliance on the power supplier in the authentication phase and to achieve mutual authentication between the gateway and the smart meter under the condition of protecting the true identity of the smart meter. In addition, the scheme can trace and revoke the identity of malicious smart meters through pseudo identity. Through the security analysis under the random oracle model and simulation tool ProVerif verifies that the proposed scheme has high security properties. Performance analysis shows that the proposed scheme can meet the requirements for security and efficiency in smart grid environment. [ABSTRACT FROM AUTHOR]

Link prediction is an important issue in the subject of data mining. The similarity algorithm based on random walk often set the probability of particles transferring to adjacent nodes to be equal, but ignore the effect of node degree on the transition probability. To solve this problem, this paper proposed a link prediction algorithm based on lowest-degree preference random walk with restart. Firstly, the algorithm redefined the transition probability of the walkers by introduce lowest-degree preference function, then applied it to the random walk with restart, and explored the effect of lowest-degree preference strategy on node similarity. The experimental results of nine real networks show that the proposed method has higher prediction accuracy, and gives more network topology information, which proves that the algorithm has certain advantages in the evaluation of node similarity. [ABSTRACT FROM AUTHOR]

Aiming at the problems of key abuse and digital certificate management in the current lattice-based proxy re-en-cryption scheme, this paper proposed a new identity-based accountable proxy re-encryption scheme by introducing the account-ability algorithm. This scheme used the user ID to calculate matrix as the public key and used the pre-image sampling algo-rithm to extract the private key, which solved the problem of digital certificate management. It used the public keys of both us-ers to calculate and generate the re-encryption key, which improved the computational efficiency of encryption and decryption. It used the proxy's public and private keys to participate in the re-encryption calculation, completed the accountability algo-rithm, and inhibited the collusion of the proxy and the delegate. The security analysis shows that the scheme satisfies the cho-sen-plaintext attack security. In terms of efficiency, the scheme has less computational complexity and ciphertext overhead. [ABSTRACT FROM AUTHOR]

Most existing vertical handoff algorithms cannot realize seamless handoff and provide stable communication services for mobile users in heterogeneous wireless networks. In order to solve the above problem, this paper proposed a vertical handoff algorithm based on mobile user location prediction to realize seamless handoff. Firstly, using the similar regular characteristics of the mobile trajectories of each user, the algorithm trained the Long Short Term Memory (LSTM) model offline and used it to predict the location of each mobile user at the next moment online, and estimated the multiattribute parameters of the future networks. After that, the algorithm used fuzzy logic to analyze the return set of all optional networks under the condition of multi-attribute parameters, and selected the network with the highest return in order to complete the handoff decision. The experimental results show that the proposed algorithm has better handover performance than the existing handover algorithms in terms of the number of handoff and the number of handoff failures under different mobile speeds and the scale of users, and is superior to other methods in terms of switching time and memory resource consumption. In conclusion, it can realize seamless handoff in heterogeneous wireless networks. [ABSTRACT FROM AUTHOR]

There is a key problem in the complete set high utility pattern mining algorithms, which is that there are too many redundant high utilityitemsets . This makes it difficult for users to find useful information in a large set of high utilityitemsets, which seriously reduces the performance of high utility pattern mining algorithms . To solve this problem, it derived a concise high utility pattern mining method, which mainly included the maximal high utility pattern, the closed high utility pattern, the top-k high utility pattern, and the combination high utility pattern mining method among the three. First, this paper introduced the related problem d escriptions of concise high utility patterns. Then, from the perspectives of candidate set generation, data structure types, and pruning strategies, the paper summarized concise high utility pattern mining methods. Finally, this paper gave further research directions for concise high utility pattern, including the processing of concise high utility pattern based on negative terms, concise high utility pattern based on time and deal with dynamic and complex data . [ABSTRACT FROM AUTHOR]

Change mining can find the possible change space in the model through the change log. However, in the scenario of information security attack-defense situation, the results of the existing change mining algorithms may not meet the security requirements, or even deviate greatly from the security standards. To solve this problem, this paper proposed a change region verification method based on game strategy, which solved the defect that the traditional change mining algorithm could not verify its effectiveness. Firstly,the algorithm constructed the benefit matrix according to the security index to calculate the Nash equilibrium point. Secondly, it verified the effectiveness of the change region of the attack-defense model through the Nash equilibrium point. And then the algorithm classified the invalid and sub effective change regions, and gave clear guidance on the response to unknown vulnerabilities. Finally,it took a typical information security event as an example to verify the feasibility of this method. [ABSTRACT FROM AUTHOR]

Granger causality discovery algorithm for discrete-time series data has important application value. The existing method smainly use Hawkes processes modeling, which cannot be applied to non-IID data and data with time-skew errors. Therefore, this paper proposed a Granger causality discovery algorithm (PTHP) for topological Hawkes processes integrating a priori constraints. Firstly, it used the constraint-based method to screen a group of causal edges with a high significance level to improve the tolerance of the algorithm to the fault time-skew errors. Then, the edges obtained in the previous step were fused into the topological Hawkes processes as a priori constraints to solve the non-IID problem between sequences. Experiments on simulted data and real-world data show the effectiveness of this method, and it won first place in PCIC 2021 causal inference competition. [ABSTRACT FROM AUTHOR]

The development of society has brought countless data, with the unbalancedness becoming a significant feature of many data sets. So it has come to be a research hotspot for machine learning on how to make those unbalanced data sets obtain better effects of classification. Based on this, this paper conducted a comprehensive research on the current unbalanced data set classification method, and made an overall interpretation and conclusion from such three aspects as the unbalanced data sampling method, the method of machine learning-based improved algorithm and the combination method. It also analyzed and took int o account many factors, including the problems solved by each method, algorithm mentality, application scenarios, as well as the advantages and disadvantages of each, and delivered a summary on potential problems of the classification methods and a prospect on the future research directions [ABSTRACT FROM AUTHOR]

In order to solve the problem that the existing super-resolution algorithm based on deep learning didn't make full use of the feature information of each level, resulting in low reconstruction accuracy and large parameter quantity, this paper proposed a double dense connection structure named densely channel attention skip connection network. In the inner structure of the network, it improved the original dense cascade block to generate a channel separable dense cascade block. The outer structure adopted a densely residual connection and attention mechanism to fuse the features extracted by the dense block to achieve the goal that less convolution layer and higher precision effect. This paper tested the network models on several benchmark datasets . The results show the proposed model has higher accuracy and fewer parameters than the other models. [ABSTRACT FROM AUTHOR]

To solve the problems that the objects with different semantic classes are identified as co-salient objects in current co-saliency detection methods, this paper proposed a CNN and semantic correction-based co-saliency detection method (CSCCD). The proposed method first adopted the guided super pixel filter to process the super pixels obtained by SLIC and the saliency results generated by DSS, which showed clear object boundaries. Then it utilized Mask R-CNN to extract semantic features. It proposed the definitions of image semantic feature and semantic consistency. It also defined the image group semantic correction to solve the problem that detected the objects with different pose belonging to a semantic class as different semantic classes. With the concept, this paper defined image group semantic correlation class, solving semantic correlation problem of multiple images. It generated the final co-saliency detection results by fusing the saliency detection regions with the image group semantic consistent regions. The experimental results on public benchmark datasets show that this algorithm can effectively highlight the whole and outline of the object, and its comprehensive performance in objective quantification is obviously improved. [ABSTRACT FROM AUTHOR]

Longest common circular subsequence (LCCS) is the longest common subsequence (LCS) of two strings with any possible circular shifting. Finding the LCS for every possible circular shifting mount is the simplest approach to solve LCCS problem, but it is very inefficient. This paper established a necessary condition for the optimum shifting amount by estimating the upper and lower bounds of the LCS length increment after a circular shifting, and thus established an iterative filtering algorithm to solve LCCS problem. This algorithm could eliminate most of the invalid candidates in several iteration according to the tests. This paper provided an approximate algorithm estimate the length of LCCS even faster, which worked with a time complexity of O (mn) . Stochastic tests show that both proposed algorithms work very well if the similarity between the given strings is sufficiently higher than that between two random strings. [ABSTRACT FROM AUTHOR]

The node-weighted Steiner tree problem is a classic NP-hard problem in the combinatorial optimization. The existing algorithms for studying the problem have the disadvantages of high time complexity and unavailability of the optimal solution. In view of the shortcomings of existing algorithms, this paper presented a backtracking algorithm based on reduced-order technique. Firstly, this paper studied the mathematical properties of the node-weighted Steiner tree problem and utilized its mathematical properties to reduce the scale of the problem. Then, it designed an upper bound sub-algorithm and a lower bound subalgorithm and utilized such sub-algorithms to prune the solution space tree of the problem and improved the search efficiency. Finally, it designed a backtracking algorithm to solve the above-mentioned problem by utilizing the upper and lower bound subalgorithms and mathematical properties. Example analysis and experimental results demonstrate that the algorithm is not only with lower time complexity, but also can obtain the optimal solution of the problem. [ABSTRACT FROM AUTHOR]

To better solve the problem of defense lag in the APT attack-defense confrontation process, and make optimal active defense decisions under the limited resources, this paper analyzed the characteristics that attacker and defender' s objectives and strategies evolve over the APT attack phase, then established an multi-stage APT-oriented attack-defense stochastic game model ( AO-ADSG) based on the non-cooperative game theory. Aiming at the fact that the utility of APT attacker and defender were unequal, this paper designed the all-assets-elements utility quantification method conforming to the APT attack features based on non-zero sum game. Analyzing the game equilibrium, it proposed the optimal active defense strategy selection algorithm. Finally, the simulation experiment of Night Dragon verifies the feasibility and correctness of the proposed method. [ABSTRACT FROM AUTHOR]

In order to solve the problem of large customer requirements ( CR) data and complex requirement correlation in high-end manufacturing industry, this paper proposed a clustering method of CR based on weighted network. This method regarded CR as the node of weighted network, regarded the correlation between CR as the edge of the weighted network, constructed the weighted network of CR, and clustered CR on the basis of weighted network with clustering algorithm, thus classified the disordered CR according to the degree of correlation. Considering the subjectivity and fuzziness of the relationship between requirements, this paper used a triangular fuzzy number to quantify the edge weights of the weighted network. Finally, this paper verified the effectiveness of the proposed method by taking the customer's requirement for tower cranes as an example. [ABSTRACT FROM AUTHOR]

To solve the problem of high-dimensional data clustering, many effective methods, such as the cascade subspace clustering( CSC) had been proposed. However, the clustering loss defined by the CSC may destroy the feature space, which leads to meaningless representative feature subspaces and this in turn hurts clustering performance. To address this issue, this paper proposed an improved algorithm that combines the AutoEncoder to preserve the data structure. Specifically, it used clustering loss as guidance to optimize the feature space and disperse data points. To constrain the manipulation and maintain the local structure between original data and their generating representations, it applied an under-complete AutoEncoder. This paper combined clustering loss and reconstruction loss to optimize the allocation of clustering labels and learn features that are suitable for clustering with local structure preservation. It adopted Adam( adaptive moment estimation) and mini-batch SGD( minibatch stochastic gradient decent) to optimize parameters for proposed algorithms. On image and text datasets, it used accuracy, normalized mutual information and adjusted rand index to prove the effectiveness and the advantages of the proposed algorithm. [ABSTRACT FROM AUTHOR]

Aiming at the problem of missing detection and false detection caused by inaccurate motion feature extraction of existing video key frame extraction algorithms, this paper proposed a reinforcement learning key frame extraction algorithm combining multi-channel feature and attention mechanism. The algorithm extracted the human skeleton joint points from the video sequence through the human posture recognition algorithm firstly. Then it used the S-GCN and ResNet50 network to extract the motion features and static features in the video sequence respectively, and performed a weighted fusion of the two. Finally, it applied the attention mechanism to calculate the importance of the video frame of the feature sequence, and used reinforcement learning to extract and optimize key frames. The experimental results show that the algorithm can solve the problem of missing and false detection in the key frame extraction of motion video. It performs well in the detection of video frames containing key actions, with high accuracy and strong stability. [ABSTRACT FROM AUTHOR]

In order to improve the convergence and diversity of the non-dominated solution set of multi-objective optimization problems, and solve the problem that the algorithm is easy to fall into the local optimum in the later stage, according to the characteristics of different differential evolution strategies, this paper added random disturbances, and proposed a decomposition multi-objective evolution algorithm with adaptive differential evolution strategy based on the improved Chebyshev mechanism (MOEA/D-ADE-levy). Firstly, it used the mixed-level orthogonal experiment to generate uniform weight vectors and applied to improve the Chebyshev mechanism to decompose the sub-problems to obtain a uniformly distributed initial population. Secondly, it divided the population into excellent individuals, intermediate individuals and poor individuals, and used different individuals. The mutation strategy used an adaptive mechanism for the mutation factor F and the crossover probability CR to improve the convergence and diversity of the non-dominated solution set. Finally, it added the Lévy random perturbation to the solution set that fell into the local optimum to increase its global search ability, and jumped out of the local optimum. The DTLZ test function was used to verify the effectiveness of the algorithm. And the proposed algorithm was compared with common algorithms such as NSGA2, NSGA3, MOEA/D, MOEA/D-DE, etc.. And the diversity and convergence analysis of the algorithm was performed using GD and IGD evaluation indicators. The results show that the algorithm has been improved and improved in terms of convergence and diversity, and can obtain a better Pareto solution set. [ABSTRACT FROM AUTHOR]

In the face of sparse data, high data dimensions and multi view clustering tasks, the traditional maximum entropy clustering algorithm will cause clustering failure because the class center tends to be consistent. In order to solve this problem, based on the traditional maximum entropy clustering algorithm, this paper introduced the class center punishment mechanism, integrated the weight matrix to achieve multi perspective division and integration, and constructed a multi view maximum entropy clustering algorithm with class center maximum. The algorithm reflected the importance of a certain perspective by adjusting the weight on each perspective, and solved the problem that the class center on each perspective tends to be consistent due to the scarcity of data and high data dimension in the multi perspective clustering task. Through a large number of experiments, it further proves that the clustering effect of this algorithm is significantly better than the traditional clustering algorithm when dealing with high-dimensional, sparse data, interference data participating and multi view data sets. [ABSTRACT FROM AUTHOR]

Accurately measuring the importance of nodes in complex networks has important guiding significance for the study of network structure and functions. Most existing node importance evaluation algorithms consider the relevant information of the node and its neighbor nodes, but ignore the influence of the topological structure between nodes on the importance of the node. To solve this problem, this paper proposed a node importance evaluation algorithm based on gravity model and relative path number. The algorithm firstly analyzed the effect of the relative shortest path number on the information dissemination between nodes, and considered the influence of factors such as non-shortest path and path distance, then defined the node importance by the sum of the interaction force between the neighbor node and the center node in the third-order range, and finally simulated it in six real networks. Experimental results show that the proposed algorithm can not only effectively distinguish the importance of different nodes in the network, but also accurately measure the importance of network nodes. [ABSTRACT FROM AUTHOR]

To solve the problems of low efficiency and poor stability of image matching with angle and scale changes, this paper proposed an improved mesh motion statistical feature point screening algorithm based on local clustering. Firstly, as seed points by local region suppression algorithm could screen feature points with high response intensity and paired appearance, and with seed points as the clustering center to obtain the minimum enclosing rectangle as the moving grid could segment images. Then, it divided the motion grid into 3 x 3 neighborhood supporting estimator grids, and it could calculate the maximum gradient value of the motion grid in eight directions as the main direction of the motion grid. Finally, it rotated the neighborhood support estimator grid of the image to be matched to the main direction of the target image motion grid, and the mesh motion statistics algorithm usually screened the matches. The experimental results show that the matching accuracy rate of the proposed algorithm is above 90% for matching images with JPEG compression transform, light change and fuzzy transform. For images with rotation and scale transformation, the matching accuracy of the proposed algorithm is about 10% higher than that of the moving grid statistic algorithm, up to more than 70%. The algorithm takes only 13 min and has high efficiency. It shows that the proposed algorithm can screen the correct matching points stably and efficiently. [ABSTRACT FROM AUTHOR]

When calculating similarity, the existing differential privacy recommendation algorithms directly calculate based on user-scheme data, but ignore the influence of scheme attributes on user preferences, fail to reflect the real preferences of users, and cannot make accurate recommendations. To solve this problem, this paper proposed differential privacy recommendation method considering user interest analysis. In this method, it collected firstly users' interest ratings for scheme attributes, and secondly clustered the user-scheme attributes rating data by K-means + +. Then, it used the differential privacy algorithm to select the nearest users, and recommended suitable schemes for the target user. Finally, taking the recommendation of nursing home schemes as an example, the experimental results show that compared with KDPC, DPCF and PNCF, the proposed algorithm can reduce the mean absolute error by about 19. 0%, 34. 0% and 37. 7% under the same privacy budget. The mean absolute error decreases by 10. 4%, 20. 3 % and 21. 4 % for the same size of the nearest neighbor set. Therefore, based on protecting the privacy of users, the algorithm further improves the recommendation accuracy. [ABSTRACT FROM AUTHOR]

Aiming at the problems of low search accuracy and slow convergence speed of traditional flower pollination algorithm (FPA) in solving complex problems, this paper proposed an improved flower pollination algorithm based on hybrid strategy (HSFPA). It adopted an adaptive transition probability strategy to improve the transition probability and dynamically balance the relationship between global pollination and local pollination. In the stage of global pollination, it proposed a dynamic global search strategy, which could not only accelerate the convergence speed of the algorithm, but also increase the diversity of pollen population and prevented pollen from falling into local optimization. The local search enhancement strategy enabled pollen to fully develop the search space around the current high-quality pollen and improved the convergence accuracy. Pollen cross-border correction strategy further strengthened the exploration ability of the algorithm. Through the simulation test of 10 benchmark functions, the experimental results show that the HSFPA algorithm has better effect in search speed and optimization accuracy. [ABSTRACT FROM AUTHOR]

To solve the problems of difficult initialization and low positioning accuracy of SLAM algorithm caused by sparse feature points and fast camera movement in practical application scenes, this paper proposed a visual-inertial SLAM algorithm. It used EDLines feature extraction and matching algorithm to introduce rich line features to improve the robustness of feature tracking. Using the complementary advantages of the inertial sensor and the visual sensor, it carried out the optimization in three steps in the initial stage through the visual-inertial initialization by steps method, so as to improve the accuracy and speed of the initialization. Experiments show that the line feature extraction and matching algorithm used in this algorithm provides faster matching speed and lower mismatching rate than the traditional LSD algorithm. Moreover, the proposed algorithm has higher accuracy and stronger robustness in the estimation of camera pose in the visual-inertial data set. [ABSTRACT FROM AUTHOR]

To improve the capability of multi-objective differential algorithm, this paper proposed a multi-objective differential evolution algorithm based on self-adaptive strategy(MODE-SS). This algorithm not only used the hyper-volume to evaluate the performance of mutation strategies and selected a suitable mutation strategy automatically, but also employed binomial and simulated binary crossover strategies to balance global and local search capabilities. The results on 10 test functions show that MODE-SS outperforms the other six state-of-the-art multi-objective optimization algorithms. Finally, this paper used MODE-SS to solve the problem of energy consumption optimization of sea-rail intermodal transportation. The results can provide a set of available solutions to decision-makers. [ABSTRACT FROM AUTHOR]

To solve the problem of low efficiency of large data clustering, this paper proposed a fast grid density clustering algorithm SGBSCAN. Firstly, this paper gave the definition of square neighborhood density clustering, and used the square neighborhood instead of the circular neighborhood to reduce the time complexity. Secondly, this paper proposed the concept of grid of square neighborhood density clustering, and determined the density relationship between core points and data points in high density region quickly. Finally, this paper proposed the grid density cluster, and used the relationship between the grid to accelerate the formation of density clusters. It applied this algorithm to 16 data sets and compared with the existing literature algorithms. The results show that the algorithm has a significant improvement in clustering efficiency. The larger the data volume, the more obvious the efficiency of the algorithm, and the algorithm is suitable for multidimensional data clustering. [ABSTRACT FROM AUTHOR]

This paper studied the problem of secure subset which most of the existing protocols that solve such problems can only keep the elements of one set private to. Under the semi-honest model, this paper constructed a secure subset protocol by using Bloom filter and Goldwasser-Micah homomorphic encryption algorithm, and proved the security of the protocol by using common simulation examples in secure multi-party computing. It used the Bloom filter to map a data set with a large number or large number field elements into a smaller data set, improved the efficiency and range of the protocol, at the same time, it used the Goldwasser-Micali homomorphic encryption algorithm to ensure the security of the protocol. Most of the relevant researches were based on the difficult problems such that secondary residuals, it is impossible to resist quantum attacks, and the secure subset computation which can resist quantum attacks is a further research direction. [ABSTRACT FROM AUTHOR]

In order to solve the problems of the MBO algorithm, such as it is easy to fall into local optimum and the convrgence speed is low, this paper proposed an improved MBO algorithm with CSMBO. Firstly, it introduced a dimension based vertical crossover operation to substitute the original migration operation of MBO, and then generated a cross migration operator. Thus, this operation could improve search ability of MBO algorithm effectively. Secondly, in order to speed up the convergence of MBO algorithm, the sharing adjustment operator with information sharing replaced the original adjustment operator. Finally, this paper utilized the greedy strategy to instead of the elite strategy of MBO, which could reduce one sorting operation and improve the calculation efficiency of MBO algorithm. To evaluate the optimization ability of this CSMBO algorithm, this paper made some experiments on a set of common benchmark functions with 30-dimensions and 50 dimensions . The results show that the proposed CSMBO algorithm has good optimization performance, and outperforms currently available three optimization approaches. [ABSTRACT FROM AUTHOR]

The uncertainty factors operation scheduling in the lead to inability to obtain accurate time of operation and ends under emergency rescue situation. In order to solve the problem, this paper proposed a grey scheduling model and a hybrid grey cuckoo algorithm. The model used three parameters interval grey number and the four parameters interval grey number to describe the uncertain time of operation and the uncertain ends. Then, it defined the possibility measure and necessity measure, proposed the index of delay confidence, which was used to estimate the probability of operation tardiness. After that, this paper established a gray mixed integer programming model with minimize average credibility of operation tardiness, and presented a solution method of the mixed grey cuckoo search algorithm. A classic example of size 6 ( 3) X 3 verified the proposed algorithm. The results show that the proposed algorithm can solve the problem very well, and the performance is better than the basic cuckoo algorithm. [ABSTRACT FROM AUTHOR]

In order to solve the problem that most algorithms are prone to lose the detailed features of the models in the case of large simplification rate, this paper introduced the concept of vertex sharpness and developed an improved folding cost based on Q EM cost, which could preserve more detailed features of the model. At the same time, aiming at the problem that most of the simplified algorithms did not contain texture processing, this paper introduced the texture variation factor to p reserve the details of the texture. On the basis of simplification, this paper proposed a mesh local optimization algorithm to solve the problem of low quality of mesh after simplification. The experimental results show that the proposed algorithm can not only p reserve the details of the model and the integrity of texture,but also obtain the model with high-quality mesh. [ABSTRACT FROM AUTHOR]

In order to improve the accuracy of truth discovery in multi-truth scene, this paper proposed a multi-ant colonies synchronization optimization based multi-truth discovery (MAC-SO-MTD) algorithm. It modeled the multi-truth discovery problem as the subset problem, which goal was maximizing the weighted sum of similarity between the set of observations provided by each data source and the set of true values of the object. On this basis, then it designed ant colony algorithm to solve the problem. It set ant colonies according to the number of objects. Based on the subset problem's structure graph, this paper used routes' probability transition equations to search for truths synchronically. After one cycle, the best route of this cycle updating and no updating were two instances of updating pheromone, which improved the convergence speed. Finally, the analysis of algorithm complexity and contrast experiment on the real data set validates the superiority of the algorithm. [ABSTRACT FROM AUTHOR]

The traditional naive Bayes classification algorithm does not divide the importance degree according to the different feature items, which makes the classification result inaccurate. In order to solve this problem, this paper introduced Jensen-Shannon(JS) divergence and used JS divergence to express the amount of information provided by the feature terms. Aiming at the deficiency of JS divergence, the paper considered from the three aspects of word frequency, text frequency and inverse category frequency corrected by coefficient of variation, and then adjusted and corrected the JS divergence. Finally, it calculated the weight of each feature and introduced the weights into the naive Bayes formula. Compared with other algorithms, it proves that this method improves the naive Bayes classification algorithm effectively. Therefore, compared with other classification algorithms, the performance of naive Bayes classification algorithm based on JS divergence feature weighting is greatly improved. [ABSTRACT FROM AUTHOR]