Aiming at the problem of low generalization of the license plate detection model， which makes it difficult to reuse in different application scenes of smart transportation， a license plate detection algorithm in unrestricted scenes based on adaptive confidence threshold was proposed. Firstly， a multi-prediction head network model was constructed， in it， the segmentation prediction head was used to reduce the model reuse pre-processing work， the adaptive confidence threshold prediction head was used to improve the model detection ability， and the multi-scale fusion mechanism and bounding box regression prediction head were used to improve the model generalization ability. Secondly， a differentiable binary network training method was adopted to learn model parameters through differentiable binary transformation combined with the training of classification confidence and confidence threshold. Finally， the Connectivity Aware Non-Maximum Suppression （CANMS） method was used to improve the post-processing speed of license plate detection， and the lightweight network ResNet18 was introduced as the backbone network of feature extraction to reduce the model parameters and further improve the detection speed. Experimental results show that in 6 scenes with different constraints in Chinese City Parking Dataset （CCPD）， the proposed algorithm can achieve the average precision of 99.5% and the recall of 99.8%， and achieves the efficient detection rate of 70 frames per second， which are better than the performance of anchor-based algorithms such as Faster Region-Conventional Neural Network （Faster R-CNN） and Single Shot MultiBox Detector （SSD）. On the three supplementary scene test sets， the license plate detection accuracy of the proposed algorithm is higher than 90% in unrestricted scenes with different resolutions， different shooting distances， and different shooting angles of pitch. Therefore， the proposed algorithm has good detection performance and generalization ability in unrestricted scenes， and can meet the requirements of model reuse.

Stock market is an essential element of financial market， therefore， the study on volatility of stock market plays a significant role in taking effective control of financial market risks and improving returns on investment. For this reason， it has attracted widespread attention from both academic circle and related industries. However， there are multiple influencing factors for stock market. Facing the multi-source and heterogeneous information in stock market， it is challenging to find how to mine and fuse multi-source and heterogeneous data of stock market efficiently. To fully explain the influence of different information and information interaction on the price changes in stock market， a graph neural network based on multi-attention mechanism was proposed to predict price fluctuation in stock market. First of all， the relationship dimension was introduced to construct heterogeneous subgraphs for the transaction data and news text of stock market， and multi-attention mechanism was adopted for fusion of the graph data. Then， the graph neural network Gated Recurrent Unit （GRU） was applied to perform graph classification. On this basis， prediction was made for the volatility of three important indexes： Shanghai Composite Index， Shanghai and Shenzhen 300 Index， Shenzhen Component Index. Experimental results show that from the perspective of heterogeneous information characteristics， compared with the transaction data of stock market， the news information of stock market has the lagged influence on stock volatility； from the perspective of heterogeneous information fusion， compared with algorithms such as Support Vector Machine （SVM）， Random Forest （RF） and Multiple Kernel k-Means （MKKM） clustering， the proposed method has the prediction accuracy improved by 17.88 percentage points， 30.00 percentage points and 38.00 percentage points respectively； at the same time， the quantitative investment simulation was performed according to the model trading strategy.

Aiming at the problem that wearable device-based fall detection cannot have good accuracy real-timely, a real-time fall detection method based on the fusion of threshold and extremely randomized tree was proposed. In this method, the wearable devices only needed to calculate the threshold value and did not need to ensure the accuracy of fall detection, which reduced the amount of calculation; at the same time, the host computer used the extremely randomized tree algorithm to ensure the accuracy of fall detection. Most of the daily actions were filtered by the wearable devices through the threshold method, so as to reduce the amount of action data detected by the host computer. In this way, the proposed method had high accuracy of fall detection in real time. In addition, in order to reduce the false positive rate of fall detection, the attitude angle sensor and the pressure sensor were integrated into the wearable devices, and the feedback mechanism was added to the host computer. When the detection result was false positive, the wrong detected sample was added to the non-fall dataset for retraining through the host computer. Through this kind of continuous learning, the model would generate an alarm model suitable for the individual. And this feedback mechanism provided a new idea for reducing the false positive rate of fall detection. Experimental results show that in 1 259 test samples, the proposed method has an average accuracy of 99.7% and the lowest false positive rate of 0.08%.

Due to the popularity of smart mobile terminals, sensitive information such as personal location privacy, check-in data privacy and trajectory privacy in the collected spatio-temporal data are easy to be leaked. In the current researches, protection technologies are proposed for the above privacy leakages respectively, and there is not a personalized spatio-temporal data privacy protection method to prevent the above privacy leakages for users. Therefore, a personalized privacy protection model for spatio-temporal data named ( p, q, ε)-anonymity and a Personalized Privacy Protection for Spatio-Temporal Data (PPP _ST) algorithm based on this model were proposed to protect the users' privacy data with personalized settings (location privacy, check-in data privacy and trajectory privacy). The heuristic rules were designed to generalize the spatio-temporal data to ensure the availability of the published data and realize the high availability of spatio-temporal data. In the comparison experiments, the data availability rate of PPP _ST algorithm is about 4.66% and 15.45% higher than those of Information Data Used through K-anonymity (IDU-K) and Personalized Clique Cloak (PCC) algorithms on average respectively. At the same time, the generalized location search technology was designed to improve the execution efficiency of the algorithm. Experiments and analysis were conducted based on real spatio-temporal data. Experimental results show that PPP _ST algorithm can effectively protect the privacy of personalized spatio-temporal data.

In order to solve the problem of feature information loss caused by the introduction of a large number of pooling layers in traditional convolutional neural networks, based on the feature of Capsule Network (CapsNet)——using vector neurons to save feature space information, a network model 3DSPNCapsNet (3D Small Pooling No dense Capsule Network) was proposed for recognizing 3D models. Using the new network structure, more representative features were extracted while the model complexity was reduced. And based on Dynamic Routing (DR) algorithm, Dynamic Routing-based algorithm with Length information (DRL) algorithm was proposed to optimize the iterative calculation process of capsule weights. Experimental results on ModelNet10 show that compared with 3DCapsNet (3D Capsule Network) and VoxNet, the proposed network achieves better recognition results, and has the average recognition accuracy on the original test set reached 95%. At the same time, the recognition ability of the network for the rotation 3D models was verified. After the rotation training set is appropriately extended, the average recognition rate of the proposed network for rotation models of different angles reaches 81%. The experimental results show that 3DSPNCapsNet has a good ability to recognize 3D models and their rotations.

Due to the large number, wide distribution and complex environments of Internet of Things (IoT) devices, IoT is more vulnerable to DDoS (Distributed Denial of Service) attacks than traditional networks. Concerning this problem, a Distributed Denial of Service (DDoS) attack detection method based on Equal Length of Value Range K-means (ELVR- Kmeans) algorithm in Software Defined IoT (SD-IoT) architecture was proposed. Firstly, the centralized control characteristic of the SD-IoT controller was used to extract the flow tables of the OpenFlow switch to analyze the DDoS attack traffic characteristics in SD-IoT environment and extract the seven-tuple features related to the DDoS attack traffic. Secondly, the obtained flow tables were classified by the ELVR- Kmeans algorithm to detect whether a DDoS attack had occurred. Finally, the simulation experiment environment was built to test the detection rate, accuracy and error rate of the method. The simulation results show that the proposed method can effectively detect DDoS attacks in SD-IoT environment with detection rate and accuracy of 96.43% and 98.71% respectively, and error rate of 1.29%.

Face recognition is widely applied in various practical conditions such as entrance guard due to its convenience and practicability. But it is vulnerable to various forms of spoofing attacks (such as photo attacks and video attacks). The liveness detection based on deep Convolution Neural Network (CNN) can solve the above problem but has disadvantages such as high calculation cost, unfriendly interaction mode and difficult deployment on embedded devices. Therefore, a real-time and lightweight security classification method of face recognition was proposed. The face liveness detection algorithm based on color and texture analysis was integrated with the face authentication algorithm, and a face recognition algorithm performing face liveness detection and face authentication in the situation of monocular camera without user cooperation was proposed. The proposed algorithm can support real-time face recognition and has higher liveness recognition rate and robustness. In order to validate the performance of the proposed algorithm, Chinese Academy of Sciences Institute of Automation-Face Anti-Spoofing Dataset (CASIA-FASD) and Replay-Attack dataset were utilized as the benchmark datasets of the experiment. The experimental results show that, in the liveness detection, the proposed algorithm has the Half Total Error Rate (HTER) of 9.7% and Equal Error Rate (EER) of 5.5% respectively, and has the time cost of 0.12 s to process a frame of image in the whole process. The above results verify the feasibility and effectiveness of the proposed algorithm.

Simultaneous Localization and Mapping (SLAM) is a key technology for robots to realize autonomous navigation in unknown environments. Aiming at the poor real-time performance and low accuracy of the commonly used RGB-Depth (RGB-D) SLAM system, a new RGB-D SLAM system was proposed to further improve the real-time performance and accuracy. Firstly, the Oriented FAST and Rotated BRIEF (ORB) algorithm was used to detect the image feature points, and the extracted feature points were processed by using the quadtree-based homogenization strategy, and the Bag of Words (BoW) was used to perform feature matching. Then, in the stage of system camera pose initial value estimation, an initial value which was closer to the optimal value was provided for back-end optimization by combining the Perspective n Point (P nP) and nonlinear optimization methods. In the back-end optimization, the Bundle Adjustment (BA) was used to optimize the initial value of the camera pose iteratively for obtaining the optimal value of the camera pose. Finally, according to the correspondence between the camera pose and the point cloud map of each frame, all the point cloud data were registered in a coordinate system to obtain the dense point cloud map of the scene, and the octree was used to compress the point cloud map recursively, so as to obtain a 3D map for robot navigation. On the TUM RGB-D dataset, the proposed RGB-D SLAM system, RGB-D SLAMv2 system and ORB-SLAM2 system were compared. Experimental results show that the proposed RGB-D SLAM system has better comprehensive performance on real-time and accuracy.

Aiming at the problem that wind turbine blade icing seriously affects the generating efficiency, safety and economy of wind turbines, a Stacked AutoEncoder (SAE) network based prediction model was proposed based on SCADA (Supervisory Control And Data Acquisition) data. The unsupervised method of encoding-decoding was utilized to pre-train the unlabeled dataset, and then the back propagation algorithm was utilized to train and fine tune the labeled dataset to achieve adaptive fault feature extraction and fault state classification. The complexy of the traditional prediction models was simplified effectively, and the influence of artificial feature extraction was avoided on model performance. The historical data of wind turbine No.15 collected by SCADA system was used for training and testing. The accuracy of the test results was 97.28%. Compared with the models based on Support Vector Machine (SVM) and Principal Component Analysis-Support Vector Machine (PCA-SVM), which accuracies are 91% and 93% respectively, the result indicates that the proposed model is more accurate than the other two.

Lack of labeled data accumulated for telecommunication fraud in the bank and high cost of manually labeling cause the insufficiency of labeled data that can be used in supervised learning methods for telecommunication fraud detection. To solve this problem, an unsupervised learning method based on dense subgraph was proposed to detect telecommunication fraud. Firstly, subgraphs with high anomaly degree in the network of accounts and resources (IP addresses and MAC addresses) were searched to identify fraud accounts. Then, a subgraph anomaly degree metric satisfying the features of telecommunication fraud was designed. Finally, a suspicious subgraph searching algorithm with resident disk, efficient memory and theory guarantee was proposed. On two synthetic datasets, the F1-scores of the proposed method are 0.921 and 0.861, which are higher than those of CrossSpot, fBox and EvilCohort algorithms while very close to those of M-Zoom algorithm (0.899 and 0.898), but the average running time and memory consumption peak of the proposed method are less than those of M-Zoom algorithm. On real-world dataset, F1-score of the proposed method is 0.550, which is higher than that of fBox and EvilCohort while very close to that of M-Zoom algorithm (0.529). Theoretical analysis and simulation results show that the proposed method can be applied to telecommunication fraud detection in the bank effectively, and is suitable for big datasets in practice.

Aiming at the problem of low network bandwidth utilization rate of the original data recovery mechanism in Information Centric Networking (ICN), a Network Coding based Real-time Data Retransmission (NC-RDR) algorithm was proposed. Firstly, the lost data packets in the network were counted according to the real-time status of the network. Then, network coding was combined into ICN, and the statistical lost data packets were combinatorially coded. Finally, the encoded data packets were retransmitted to the receiver. The simulation results show that compared with NC-MDR (Network Coding based Multicast Data Recovery) algorithm, in the transmission bandwidth aspect, the average number of transmissions was reduced by about 30%. In ICN, the proposed algorithm can effectively reduce the number of data re-transmissions, improveing network transmission efficiency.

Fuzzy C-Means (FCM) clustering algorithm is sensitive to the initial clustering center and is easy to fall into local optimum. Therefore, a Firefly Fuzzy C-Means clustering Algorithm based on Levy flight (LFAFCM) was proposed. In LFAFCM, the random movement strategy of firefly algorithm was changed to balance the algorithm's local search and global search capabilities, the Levy flight mechanism was introduced during the firefly position update process to improve the global optimization ability, and the scale coefficient of each firefly was dynamically adjusted according to the number of iterations and the firefly position to limit the searchable range of Levy flight and speed up the convergence of the algorithm. The algorithm was validated by using five UCI datasets. The experimental results show that the algorithm avoids the local optimum and has a fast convergence speed.

Focused on the issue that Point-Of-Interest (POI) recommendation has low recommendation accuracy and efficiency, with deep analysis of the influence of social factors and geographical factors in POI recommendation, a POI recommendation algorithm combining location influence was presented. Firstly, in order to solve the sparseness of sign-in data, the 2-degree friends were introduced into the collaborative filtering algorithm to construct a social influence model, and the social influence of the 2-degree friends on the users were obtained by calculating experience and friend similarity. Secondly, by deep consideration of the influence of geographical factors on POI, a location influence model was constructed based on the analysis of social networks. The users' influences were discovered through the PageRank algorithm, and the location influences were calculated by the POI sign-in frequency, obtaining overall geographical preference. Moreover, kernel density estimation method was used to model the users' sign-in behaviors and obtain the personalized geographical features. Finally, the social model and the geographic model were combined to improve the recommendation accuracy, and the recommendation efficiency was improved by constructing the candidate POI recommendation set. Experiments on Gowalla and Yelp sign-in datasets show that the proposed algorithm can quickly recommend POIs for users, and has high accuracy and recall rate than Location Recommendation with Temporal effects (LRT) algorithm and iGSLR (Personalized Geo-Social Location Recommendation) algorithm.

In view of the problem that the existing researches do not consider the refinement of the factors affecting the network negative energy propagation and construct a propagation dynamics model for analysis, a Weak-Strong-Received-Infected-Evil (WSRIE) model of network negative energy propagation was proposed. Firstly, considering the difference of negative energy immunity and propagation ability of network users, the vulnerable states were divided into weak immunity and strong immunity, and the infectious states were divided into weak infection, strong infection and malicious propagation with unchanged scale. Secondly, according to the negative energy infection mechanism of the network, the state transition law was proposed. Finally, a dynamics model of network negative energy propagation for complex networks was constructed. The simulation comparison experiments on NW small world network and BA scale-free network were carried out. The simulation results show that under the same parameters, the weak immune node density of the NW network is 9 percentage points lower than that of the BA network, indicating that the network with small world characteristics is more susceptible to negative energy. In the BA network, the density of infected nodes with the malicious node degree of 200 is 5 percentage points higher than that with the node degree of 0, indicating that the greater the node degree of the network red opinion leader, the more network users affected by the network negative energy.

To solve the sparse and cold start problems of recommendation system, a Trust Regularization Euclidean Embedding (TREE) algorithm by fusing trust information was proposed. Firstly, the Euclidean embedding model was employed to embed the user and project in the unified low-dimensional space. Secondly, to measure the trust information, both the project participation degree and user common scoring factor were brought into the user similarity calculation formula. Finally, a regularization term of social trust relationship was added to the Euclidean embedding model, and trust users with different preferences were used to constrain the location vectors of users and generate the recommendation results. In the experiments, the proposed TREE algorithm was compared with the Probabilistic Matrix Factorization (PMF), Social Regularization (SoReg), Social Matrix Factorization (SocialMF) and Recommend with Social Trust Ensemble (RSTE) algorithms. When dimensions are 5 and 10, TREE algorithm has the Root Mean Squared Error (RMSE) decreased by 1.60% and 5.03% respectively compared with the optimal algorithm RSTE on the dataset Filmtrust.While on the dataset Epinions, the RMSE of TREE algorithm was respectively 1.12% and 1.29% lower than that of the optimal algorithm SocialMF. Experimental results show that TREE algorithm further alleviate the sparse and cold start problems and improves the accuracy of scoring prediction.

As the core of the operation and management of a rail transit enterprise, the rail transit operation organization plays a very important role in reducing the operation cost of the enterprise, improving the service level and the travel efficiency of passengers. A strategy based on Artificial Bee Colony (ABC) optimization algorithm was proposed to optimize the train traffic interval. Based on the consideration of the respective interests of operators and passengers, the train departure interval was taken as the decision variable to establish a bi-objective nonlinear programming model for the lowest average passenger waiting time and the largest train waiting time. Artificial Bee Colony (ABC) algorithm was used to optimize the model. The simulation results on Beijing-Tianjin inter-city passenger flow at different times of a day demonstrate the effectiveness of the proposed algorithms and models.

It is difficult to realize on-line measurement for the large dimension range of China Railway High-speed (CRH) body, the complexity of testing items and the variety of vehicles. Firstly, a measurement scheme of key dimensions for a large-scale bullet train was proposed, where binocular Charge Coupled Device (CCD) stereo vision was used to set up the measuring sub stations of each key dimension, and the laser tracker and coordinate transformation algorithm were used to complete the global calibration of each CCD camera's measuring sub station. In each measuring sub station, the stereo spatial ball detection technology was used to measure local key dimensions. At the same time, a neural network temperature error compensation model based on wavelet analysis was constructed, and the precision of space distance compensation reached 0.05 mm. The comparison between the proposed method and three-coordinate measuring machine, shows that the proposed method is simple in operation, high in flexibility and high in precision, which can effectively solve the key dimension detection problem of CRH body.

To solve problems such as easily getting trapped in local optimum and slow convergence rate in BareBones Particle Swarm Optimization (BBPSO) algorithm, a dynamic Multi-Subgroup collaboration Barebones Particle Swarm Optimization based on Kernel Fuzzy Clustering (KFC-MSBPSO) was proposed. Based on the standard BBPSO algorithm, firstly, kernel fuzzy clustering method was used to divide the main group into several subgroups, and the subgroups optimized collaboratively to improve the searching efficiency. Then, nonlinear dynamic mutation factor was introduced to control subgroup mutation probabilities according to the number of particles and convergence conditions, the main group was reconstructed by means of particle mutation and the exploration ability was improved. The main group particle absorption strategy and subgroup merge strategy were proposed to strengthen the information exchange between main group and subgroups and enhanced the stability of the algorithm. Finally, the subgroup reconstruction strategy was used to adjust the iterations of subgroup reconstruction by combining the optimal solutions. The results of experiments on six benchmark functions, such as Sphere, show that the accuracy of KFC-MSBPSO algorithm has improved by at least 11.1% compared with classical BBPSO algorithm, Opposition-Based Barebones Particle Swarm Optimization (OBBPSO) algorithm and other improved algorithms. The best mean value in high dimensional space accounts for 83.33% and has a faster convergence rate. This indicates that KFC-MSBPSO algorithm has good search performance and robustness, which can be applied to the optimization of high dimensional complex functions.

For the problem that qualitative analysis methods are mostly adopted for the study of spam information propagation and it is difficult to reveal intrinsic propagation rules of spam information propagation, an ILDR (Ignorant-Lurker-Disseminator-Removed) model of spam information was proposed based on the idea of virus propagation modeling by considering the actual factors such as different input rates and removal rates. Firstly, the equilibrium point and the propagation threshold were calculated, and the stability conditions of the equilibrium point were given. Secondly, the local stability of non-spam information and spam information was proved by the Routh-Hurwitz criterion, then the global stability of non-spam information was certified via the invariance principle of LaSlle, which was proved based on the Bendixson criterion. Theoretical research shows that the non-spam information equilibrium is global asymptotically stable when propagation threshold is less than 1; the spam information equilibrium is global asymptotically stable when propagation threshold is greater than 1. The numerical simulation validates that the value of the propagation threshold can be decreased when decreasing the transfer rate from the lurker to the disseminator, increasing the transfer rate from the ignorant to the remover and the transfer rate from the lurker to the remover; the value of the disseminator can be decreased via increasing the proportionality coefficient from the ignorant to the lurker, or increasing the transfer rate of the disseminator to the remover and the system removal rate.

To solve the problem of two non-smooth regularization terms in sparse reconstruction of Magnetic Resonance Imaging (MRI) based on Compressed Sensing (CS), a new Proximal Smoothing Iterative Algorithm (PSIA) based on Moreau-envelope was proposed. The classical sparse reconstruction for MRI based on CS is a problem of minimizing the objective function with a linear combination of three terms:the least square data fidelity term, the sparse regularization term of wavelet transform, and the Total Variation (TV) regularization term. Firstly, the proximal smoothing of the wavelet transform regularization term in the objective function was carried out. Then, the linear combination of the data fidelity term and the wavelet transform regularization term after smooth approximation was considered as a new convex function that could be continuously derived. Finally, PSIA was used to solve the new optimization problem. The proposed algorithm can not only cope with the two regularization constraints simultaneously in the optimization problem, but also avoid the algorithm robustness problem caused by fixed weights. The experimental results on simulated phantom images and real MR images show that, compared with four classical sparse reconstruction algorithms such as Conjugate Gradient (CG) decent algorithm, TV l₁ Compressed MRI (TVCMRI) algorithm, Reconstruction From Partial k space algorithm (RecPF) and Fast Composite Smoothing Algorithm (FCSA), the proposed algorithm has better reconstruction results of image signal-to-noise ratio, relative error and structural similarity index, and its algorithm complexity is comparable to the existing fastest reconstruction algorithm FCSA.

Aiming at the current status of Niederreiter public key cryptosystem which is vulnerable to distinguishing attack and ISD (Information Set Decoding), an improved Niederreiter public key cryptosystem was proposed. Firstly, the permutation matrix in the Niederreiter cryptography scheme was improved, and the original permutation matrix was replaced by a random matrix. Secondly, the error vector in the Niederreiter cryptography scheme was randomly divided to conceal the Hamming weight. Finally, the encryption and decryption processes of the Niederreiter cryptography scheme were improved to improve the security. The analysis shows that the improved scheme can resist the distinguishing attack and ISD. The public key size of the improved scheme is smaller than that of the scheme proposed by Baldi, et al. (BALDI M, BIANCHI M, CHIARALUCE F, et al. Enhanced public key security for the McEliece cryptosystem. Journal of Cryptology, 2016, 29(1):1-27). At the 80-bit security level, the public key of the improved scheme is reduced from 28408 bits to 4800 bits. At the 128-bit security level, the public key size of the improved scheme is reduced from 57368 bits to 12240 bits. As one of the anti-quantum cryptography schemes, the viability and competitiveness of the improved scheme are enhanced.

Coding-based encryption scheme, with the advantages of anti-quantum feature and fast encryption and decryption speed, is one of the candidate schemes for anti-quantum cryptography. The existing coding-based hybrid encryption schemes have the INDistinguishability under Chosen Ciphertext Attack (IND-CCA) security, which have the disadvantage that the public key size used to encrypt the shared secret key of the sender and receiver is large. The problem of large size of public key in hybrid encryption scheme based on Niederreiter coding was solved by the following three steps. Firstly, the private key of Niederreiter coding scheme was randomly split. Then, the plaintext of Niederreiter coding scheme was split randomly. Finally, the encryption and decryption processes of Niederreiter coding scheme were improved. It is concluded through analysis that, the public key size of the improved scheme is less than that of Maurich scheme. Compared with Maurich scheme, the public key of the improved scheme is reduced from 4801 bits of the original scheme to 240 bits under the security level of 80 bits, and the public key of the improved scheme is reduced from 9857 bits to 384 bits under the security level of 128 bits. Although the improved scheme is more complicated than the original scheme, its storage cost and calculation cost are smaller, and the practicability of the improved scheme is enhanced.

In recent years, the amount of data in various industries grows rapidly, which results in the increasing of optimization demands in database storage system. Relational databases are I/O-intensive, take use of relatively free CPU time, data compression technology could save data storage space and I/O bandwidth. However, the compression features of current database systems were designed for traditional storage and computing environments, without considering the impact of virtualized environments or the use of Solid State Drive (SSD) on system performance. To optimize the cache performance of database compression system, a database compression system performance model was proposed, and the impact on the I/O performance of various system environments was analyzed. Take the open source database MySQL as an example, the corresponding cache optimization methods were given based on analysis. Evaluation results on Kernel-based Virtual Machine (KVM) and MySQL database show that the optimized version has an increase of more than 40% in performance under some configurations, even close to superior physical machine performance.

The clustering algorithm of Gaussian Mixture Model (GMM) is sensitive to initial value and easy to fall into local minimum. In order to solve the problems, taking advantage of strong global search ability of Density Peaks (DP) algorithm, the initial clustering center of GMM algorithm was optimized, and a new Clustering algorithm of GMM based on DP (DP-GMMC) was proposed. Firstly, the clustering center was searched by the DP algorithm to obtain the initial parameters of mixed model. Then, the Expectation Maximization (EM) algorithm was used to estimate the parameters of mixed model iteratively. Finally, the data points were clustered according to the Bayesian posterior probability criterion. In the Iris data set, the problem of dependence on the initial clustering center is solved, and the clustering accuracy of DP-GMMC can reach 96.67%, which is 33.6 percentage points higher than that of the traditional GMM algorithm. The experimental results show that, the proposd DP-GMMC has better clustering effect on low-dimensional datasets.

The impact of individual interest and refutation mechanism on rumor spread was investigated, and a new IWSR (Ignorant-Weak spreader-Strong spreader-Removal) rumor spread model was proposed. The basic reproduction number and equilibrium points of the model were calculated. Using Lyapunov stability theorem, Hurwitz criterion and LaSalle invariance principle, the local stability and global stability of some equilibrium points were proved. Through numerical simulations, it is concluded that increasing the effectivity of the governments' refutation actions or improving people' ability of rumor judgement can effectively suppress rumor spread. Finally, numerical simulations were conducted in WS (Watts-Strogatz) small-world network and BA (Barabási-Albert) scale-free network, showing that the network topology exerts significant influence on rumor spread.

Due to a large amount of data and high dimension in currently network intrusion detection, a Moth-Flame Optimization (MFO) algorithm was applied to the feature selection of network intrusion detection. Since MFO algorithm converges fast and easy falls into local optimum, a Binary Moth-Flame Optimization integrated with Particle Swarm Optimization (BPMFO) algorithm was proposed. On one side, the spiral flight formula of the MFO algorithm was introduced to obtain strong local search ability. On the other side, the speed updating formula of the Particle Swarm Optimization (PSO) algorithm was combined to make the individual to move in the direction of global optimal solution and historical optimal solution, in order to increase the global convergence and avoid to fall into local optimum. By adopting KDD CUP 99 data set as the experimental basis, using three classifiers of Support Vector Machine (SVM), K-Nearest Neighbor ( KNN) and Naive Bayesian Classifier (NBC), Binary Moth-Flame Optimization (BMFO), Binary Particle Swarm Optimization (BPSO), Binary Genetic Algorithm (BGA), Binary Grey Wolf Optimization (BGWO) and Binary Cuckoo Search (BCS) were compared in the experiment. The experimental results show that, BPMFO algorithm has obvious advantages in the comprehensive performance including algorithm accuracy, operation efficiency, stability, convergence speed and jumping out of local optima when it is applied to the feature selection of network intrusion detection.

To solve the problem that root branching plasticity cannot be achieved by using sequential model in root branch modeling, a new root branching method based on swarm Parrondo's game was proposed to analyze root branching plasticity in heterogeneous root growth environment. Firstly, root primordial swarm based on individual root primordium was constructed. Secondly, Parrondo's game was used to achieve interaction among root primordial swarm affected by environment. Finally, root branch modeling process was simulated according to auxin that was updated based on the interaction results of root primordial. Prediction of root branching probability was achieved in four different root growth environments. The experimental results show that compared with RootMap, etc, the proposed method can be used to model development process of root primordium into root branch affected by spatial and temporal changes in the growth environment of the root system, and also provides analysis means for root system modeling and simulation research.

Focusing on the linear decomposition of the S-box layer in Keccak algorithm, a new linear structure construction method was proposed based on the algebraic properties of the S-box. Firstly, to ensure the state data was still linear with that after this linear structure, some constraints about input bits of S-box needed to be fixed. Then, as an application of this technique, some new zero-sum distinguishers of round-reduced Keccak were constructed by combining the idea of meet-in-the-middle attack. The results show that a new 15-round distinguisher of Keccak is found, which extends 1-round forward and 1-round backward. This work is consistent with the best known ones and its complexity is reduced to 2 ²⁵⁷. The new distinguisher, which extends 1-round forward and 2-round backward, has the advantages of more free variables and richer distinging attack combinations.

Aimming at the problem of leaking privacy to Third-Party Auditors (TPA) and initiating alternative attacks by Cloud Storage Server (CSS) in public auditing, a new public auditing scheme of data integrity for public cloud was proposed. Firstly, the hash value obfuscation method was used to obfuscate the evidence returned by the cloud storage server to prevent TPA from analyzing and calculating the original data. Then during the audit process, TPA itself calculated the overlay tree of the Merkle Hash Tree (MHT) corresponding to the challenge request, and matched with the overlay tree returned by CSS to prevent the cloud storage server from responding to audit challenges with other existing data. Experimental results show that the performance in terms of computational overhead, storage overhead and communication overhead does not change by orders of magnitude after solving the privacy and attack problems of the existing scheme.

Considering the problem that traditional active learning algorithms can only handle small and medium size data sets, a big data active learning algorithm based on MapReduce was proposed. Firstly, a classifier was trained by Extreme Learning Machine (ELM) algorithm on an initial training set, and the outputs of the classifier were transformed into a posterior probability distribution by softmax function. Secondly, the big data set without labels was partitioned into l subsets, which were deployed to a cloud computing platform with l nodes. On each node, the information entropies of instances of each subset were calculated by the trained classifier, and q instances with maximum information entropies were selected for labeling, then the l× q labeled instances were added into the training set. Repeat the above steps until the predefined termination criterion was satisfied. Contrast test with ELM-based active learning algorithm were conducted on 4 data sets including Artificial, Skin, Statlog and Poker. Experimental results show that the proposed algorithm can complete active instance selection on 4 data sets, while the active learning algorithm based on ELM can only complete active instance selection on the smallest data set, indicating that the proposed algorithm outperforms the active learning algorithm based on ELM.