The problem of network traffic classification has always been a challenge of iterative methods with the development of network communication， and many solutions have been developed. At present， most network data classification methods focus on the balanced dataset to facilitate experiment and calculation. To solve the problem that most real network datasets are still unbalanced， a network traffic classification model VAE-ABC （Variational AutoEncoder- Adaptive Boosting-Convolutional neural network） was proposed by integrating Variational AutoEncoder （VAE） and Adaptive Boosting Convolutional Neural Network （AdaBoost-CNN）. Firstly， at the data level， VAE was used to partially enhance the unbalanced dataset， and shorten the learning time with the VAE’s characteristics of learning data potential distribution. Then， in order to improve classification effect at the algorithm level， combining with the idea of ensemble learning， AdaBoost-CNN algorithm was designed on the basis of Adaptive Boosting （AdaBoost） algorithm with using an improved Convolutional Neural Network （CNN） as a weak classifier， thereby improving the accuracy of learning and training. Finally， the fully connected layer was used to complete feature mapping， and then the final classification results were obtained through an activation function Sigmoid. After multiple comparisons， experimental results show that the proposed model achieves an accuracy of 94.31% on the unbalanced sub-dataset of partitioned classification dataset ISCX VPN-nonVPN. Compared with AdaBoost-SVM， using Support Vector Machine （SVM） as a weak classifier， SMOTE-SVM， combining SMOTE （Synthetic Minority Oversampling TEchnique） and SVM， and SMOTE-AB-D-T， with Decision Tree （D-T） as a weak classifier and combined with SMOTE algorithm， the proposed model has the accuracy increased by 1.34， 0.63 and 0.24 percentage points， respectively. It can be seen that the classification effect of this model is better than those of other models on this dataset.

In recent years， deep learning has made breakthroughs in many fields due to its powerful representation capability， and the architecture of neural network is crucial to the final performance. However， the design of high-performance neural network architecture heavily relies on the priori knowledge and experience of the researchers. Because there are a lot of parameters for neural networks， it is difficult to design optimal neural network architecture. Therefore， automated Neural Architecture Search （NAS） gains significant attention. NAS is a technique that uses machine learning to automatically search for optimal network architecture without the need for a lot of human effort， and is an important means of future neural network design. NAS is essentially a search optimization problem， by designing search space， search strategy and performance evaluation strategy， NAS can automatically search the optimal network structure. Detailed and comprehensive analysis， comparison and summary for the latest research progress of NAS were provided from three aspects： search space， search strategy， and performance evaluation strategy， which facilitates readers to quickly understand the development process of NAS. And the future research directions of NAS were proposed.

Focusing on the issue that embedding the attention mechanism module into Convolutional Neural Network （CNN） to improve the application accuracy will increase the parameters and the computational cost， the lightweight Height Dimensional Squeeze and Excitation （HD-SE） module and Width Dimensional Squeeze and Excitation （WD-SE） module based on squeeze and excitation were proposed. To make full use of the potential information in the feature maps， two kinds of height and width dimensional weight information of feature maps was respectively extracted by HD-SE and WD-SE through squeeze and excitation operations， then the obtained weight information was respectively applied to corresponding tensors of the feature maps of two dimensions to improve the application accuracy of the model. Experiments were implemented on CIFAR10 and CIFAR100 datasets after embedding HD-SE and WD-SE into Visual Geometry Group 16 （VGG16）， Residual Network 56 （ResNet56）， MobileNetV1 and MobileNetV2 models respectively. Experimental results show fewer parameters and computational cost added by HD-SE and WD-SE to the network models when the models achieve the same or even better accuracy， compared with the state-of-the-art attention mechanism modules， such as Squeeze and Excitation （SE） module， Coordinate Attention （CA） block， Convolutional Block Attention Module （CBAM） and Efficient Channel Attention （ECA） module.

In order to solve the problems such as high time cost， inaccuracy and influence of parameter setting on algorithm performance when optimizing parameters of Convolutional Neural Network （CNN） by traditional manual methods， a variable Convolutional AutoEncoder （CAE） method based on Teaching-Learning-Based Optimization （TLBO） was proposed. In the algorithm， a variable-length individual encoding strategy was designed to quickly construct the CAE structure， and stack CAEs to a CNN. In addition， the excellent individual structure information was fully utilized to guide the algorithm to search the regions with more possibility， thereby improving the algorithm performance. Experimental results show that the classification accuracy of the proposed algorithm achieves 89.84% when solving medical image classification problems， which is higher than those of traditional CNN and similar neural networks. The proposed algorithm solves the medical image classification problems by optimizing the CAE structure and stacking CNN， and effectively improves the classification accuracy of medical image classification.

Aiming at the problem of the design of Deoxyribonucleic Acid (DNA) encoding sequences which can be mathematically converted into a multi-objective optimization problem with some constraints, by introducing the h-distance in the set of DNA single strands, a sharing function between different DNA sequences was defined and a micro-genetic algorithm was applied to solve the DNA encoding sequence problem. Compared with the previous results, the algorithm can get better DNA sequences and improve the efficiency of computation. The algorithm can be used to design concrete DNA sequences in DNA computing.

The popular clustering ensemble algorithms cannot give the appropriate treatment program in the light of the different characteristics of the different data sets. A new clustering ensemble algorithm — Enhanced Clustering Ensemble algorithm based on Characteristics of Data sets (ECECD) was proposed for overcoming this defect. ECECD was composed of generation of base clustering, selection of base clustering and consensus function. It selected a special range of ensemble members to form the final ensemble and produced the final clustering based on the characteristic of the data set. Three Benchmark data sets including ecoli, leukaemia and Vehicle were clustered in the experiment, and the clustering errors gained by the proposed algorithm were 0.014, 0.489 and 0.361 respectively, which were always the minimum compared with that of the other algorithms such as Bagging based Structure Ensemble Approach (BSEA), Hybrid Cluster Ensemble (HCE) and Cluster-Oriented Ensemble Classifier (COES). The Normalized Mutual Information (NMI) values of the proposed algorithm were also always higher than that of these algorithms when increasing candidate base clusterings. Therefore, compared with these popular clustering ensemble algorithms, the proposed algorithm has the highest clustering precision and the strongest scalability.

Kernel Principal Component Analysis (KPCA) and Multi-Layer Perceptron (MLP) neural network are popular feature extraction algorithms. However, these algorithms are inefficient and easy to fall into local optimal solution. The paper proposed a new feature extraction algorithm — margin maximizing hyperplanes based Enhanced Feature Extraction algorithm （EFE）, which can overcome the problem of KPCA and MLP algorithm. The proposed EFE algorithm, whcih maps the input samples to the subspace spanned by the normals of hyperplanes through adopting the pairwise orthogonal margin maximizing hyperplanes, is independent of the probability distribution of the input samples. The results of these feature extraction experiments on real world data set — wine and AR show that FE algorithm is beyond KPCA and MLP in terms of the efficiency of the implementation and accuracy of recognition.

The (t,n) threshold ring signature could be generated by any t entities of n entities group on behalf of the whole group, while the actual signers remain anonymous. In order to design the threshold ring signature scheme with constant size, this paper presented an identity-based threshold ring signature scheme without random oracle by using bilinear pairing technique. In the end, the authors prove this scheme satisfy the unconditional signer ambiguity and existential unforgeability against selective identity, selective chosen message attack in terms of the hardness of Diffie-Hellman Inversion （DHI） problem.

Humanoid information grid management model(HIGMM) based on generalized artificial life was proposed after researching the principle and approaches of generalized artificial life and information grid.The design idea of information grid management model with such humanoid management characteristics as dual management scheme,central-decentralized management pattern and multi-level coordination function was discussed.

Based on the implements of the secure mobile auction system, the paper analysed the security requirements of this system, introduced the special encryption and sign techniques in mobile auction system, these techniques contained the making, storing and updating of the secret key, the encryption, decryption and the self-sign certificate model of SOAP message in server, the implement of the encryption engine and the sign engine in client. This system can be successfully transplanted to mobile devices in order to improve the security of the wireless Web Services.