In the protein-protein interactions, hub proteins are the key factor to maintain stability of protein-protein interactions and exert the protein biological function. Conventional methods mainly focus on the topological structure and gene expression of hub proteins, while we mainly discuss the hot spots of hub protein interfaces. In order to evaluate the performance of the classification models, the importance of feature variables is analyzed by using the average precision descent curve and the average Gini coefficient descent curve. In addition, the margin box-plot is used to measure the certainty of the classification models. The experimental results show that the error rate of random forest method is lower, and our classification model has higher reliability.

Protein structure prediction is an important factor in the area of bioinformatics. Predicting the two-dimensional structure of proteins based on the hydrophobic polarity model (HP model) is a typical non-deterministic polynomial(NP)-hard problem. Currently, HP model optimization methods include the greedy algorithm, particle swarm optimization, genetic algorithm, ant colony algorithm and the Monte-Carlo simulation method. However, the robustness of these methods are not sufficient, and it is easy to fall into a local optimum. Therefore, a HP model optimization method, based on reinforcement learning was proposed. In the full state space, a reward function based on energy function was designed and a rigid overlap detection rule was introduced. By using the characteristics of the continuous Markov optimal decision and maximizing global cumulative return, the global evolutionary relationship in biological sequences was fully exploited, and effective and stable predictions were retrieved. Eight classical sequences from publications and Uniref50 were selected as experimental objects. The robustness, convergence and running time were compared with the greedy algorithm and particle swarm optimization algorithm, respectively. Both reinforcement method and swarm optimization method can find all the lowest energy structures for these eight sequences, while the greedy algorithm only detects 62.5%. Compared with particle swarm optimization, the running time of the reinforcement method is 63.9% lower than that of particle swarm optimization.

To address the flaws in clustering speed, this paper proposes a novel PSO-GGA clustering algorithm based on pattern reduction. To fully combine the pattern reduction method, the algorithm uses a generalized genetic algorithm in serial to improve the particle swarm optimization algorithm. This can increase the diversity of samples and protect patterns that need to be saved for compression. At the same time, to determine the number of particles needed to replace the poor particles an incremental strategy is employed. This fully embodies the PSO’s ability for rapid search optimization and the genetic algorithm’s advantage of a large search space. The experimental results show that the clustering time only required 20% compared to the original algorithm without showing any obvious decline in accuracy.

In order to overcome the premature convergence of particle swarm optimization (PSO) algorithm, an improved PSO algorithm based on sub-groups mutation (SsMPSO) is proposed. This algorithm has proposed the sub-groups with random directional vibrating search to mutate the global optimal position of the main swarm and changed the way of random mutation. The mutation based on sub-groups enabled the algorithm had excellent local exploit ability and circumvented the premature convergence. It used another mutation on bad particles to enhance the algorithm's global exploit ability and expand the searching space. Finally, high dimension benchmark functions have been used to test the performance of improved algorithm. The simulation results show that the proposed algorithm can effectively overcome the premature problem, the multimodal function optimization can avoid local extreme point and the convergence and convergence accuracy are greatly improved.

通过比较研究,探索利用现有的媒介内容、在线内容和基于AR的学习内容来帮助学生进行大学课程学习的有效性.为了在比较过程中确认学习效果,进行了准实验设计研究.通过调查学生对使用基于AR技术的学习策略和在线学习来支持学习的看法,以及基于AR技术和在线学习的学习内容如何影响学生的学习动机和表现,分析AR技术的教学应用情况.

The study of protein-protein interactions and protein structure through computational methods is critical to understand protein function. Hot spot residues play an important role in bioinformatics to reveal life activities. However, conventional hot spots prediction methods may face great challenges. This paper proposes a hot spot prediction method based on feature selection method SVM-RFE to improve the training performance. SMOTE based oversampling is used to adds new samples to avoid an overfitting classifier. SVM-RFE is then invoked to obtained optimal feature subset. Finally, a feature-based SVM is created to predict the hot spots. Experimental results indicate that the performance of hot spots prediction has been significantly improved compared with the previous methods.

为了探讨增强现实(AR)技术对于英语教学的影响,本研究选取我校某专业40名大学生,在其英语学习过程中使用基于增强现实技术的学习软件.研究数据证明增强现实技术可以让学生更积极主动学习英语,旨在为增强现实技术引入英语教学提供理论支持.

To keep the balance between the global search and local search,a new binary quantum behavior particle swarm optimization algorithm (CCBQPSO) is proposed by introducing a comprehensive learning and cooperative method in the binary QPSO algorithm.The complete learning strategy can keep the diversity of the population,and the cooperative method can directly introduce the algorithm into the local search and converge to the optimal solution quickly.In this algorithm,the individual optimal position of all particles can first participate in the updating of the local attractor;and the new solution vector dimension of each particle will replace the previous optimal individual position of the corresponding particle and the global optimal position of the population and calculate the fitness value.The results show that the proposed algorithm can increase diversity of swarm and converge more rapidly than other binary algorithms.

The experiment of alanine scanning has shown that most of the binding energies in protein-protein interactions are contributed by a few significant residues at the protein-protein interfaces, and those important residues are called hot spot residues. On the basis of protein-protein interaction, hot spot residues tend to get together to form modules, and those modules are defined as hot regions. So, hot spot residues play an important role in revealing the life activities of organisms. Therefore, how to predict hot spot residues and non-spot residues effectively and accurately is a vital research direction. A new method is proposed combining protein amino acid physicochemical features and structural features to predict the hot spot residues based on the ensemble learning. The experimental results demonstrate that this method of prediction hot spot residues has a good effect.

As the representative technology of protein spatial structure exploration, NMR technology provides an unprecedented opportunity for modern life science research. But subsequent large data analysis has become a major problem. It is an important means to study protein structure and functional relationship by known information proteins’ three-dimensional structures to predict the unknown spatial structure of proteins. A method for similarity comparison of 3D protein structures based on Riemannian manifold theory is proposed in this paper. By constructing Cα frames and extracting geometric feature of protein, 3D coordinates of proteins are converted into one dimension sequences with rotation and translation invariance. The Riemann distance is used as the three-dimensional structure similarity degree index. Spatial transformation on protein structure is not needed in this method, which avoiding errors when matching two proteins in the traditional method for registration by the least squares fitting. This method is independent of sequence information completely. It has realistic significance for proteins which do not have a similarity between sequences. Three experiments are designed according to 3 sets of data: proteins of different similarity, ten pairs whose protein structures are more difficult to identify proposed by Fischer, 700 proteins in the HOMSTRAD database. Compared with the traditional method, the experiment results show that the matching accuracy of this method has been greatly enhanced.

提出一种新的基于水平集的图像分割方法,通过引入贝叶斯规则,设计一个可有效分割弱边缘的非线性自适应速度和概率加权停止函数.该方法包含如下特性:可以自动决定曲线收缩或利用贝叶斯规则对涉及到的图像区域特征进行扩展;以恰当的速度驱动曲线演变,避免了弱边缘的遗漏;降低了假边缘的影响.最后将所提出的分割方法应用于人工图像、医学图像和自然图像的定性和定量评估.对结果进行比较可知,该方法相对于水平集方法和其代表的变体更加有效和实用.

针对传统超分辨率重建方法计算复杂度高、重建效果差等问题,提出一种基于稀疏表示的图像超分辨率重建模型.该模型利用稀疏表示方法,结合自回归原理将原始图像表示为若干个图像块的线性组合,并根据图像边缘特征将图像块进行划分,以提高算法效率,最后结合分治思想、变量分离技术以及增广拉格朗日方法对模型进行求解.实验结果表明,与传统插值算法相比,该算法对图像重建效果更好.

Protein structure prediction is defined as predicting the tertiary structure from the primary structure of the protein sequence. Because the real protein structure is very complex, it is necessary to adopt the simplified structure model for studying protein 3D space structure. In this paper, we introduce a kind of 3D AB off-lattice model for protein structure prediction, and the amino acids are labeled as two hydrophobic amino acids and hydrophilic amino acids. When the protein model is simplified, the optimization algorithm is also needed to use for searching the lowest energy conformation of the protein sequence based on the hypothesis theory. In this paper, a hybrid algorithm which combines PSO algorithm based on local adjust strategy (LAPSO) and genetic algorithm, was proposed to search the space structure of the protein with AB off-lattice model. Experimental results show that the minimal energy values obtained by the improved LAPSO are lower than those obtained by previous methods. The performance of our improved algorithm is better, and it can effectively solve the search problem of the protein space folding structure.

To solve the non-convex and non-linear economic dispatch problem efficiently, a chaotic iteration particle swarm optimization algorithm is presented. In the global research of particle swarm optimization and local optimum, ergodicity of chaos can effectively restrain premature. To balance the exploration and exploitation abilities and avoid being trapped into local optimal, a new index, called iteration best, is incorporated into particle swarm optimization, and chaotic mutation with a new Tent map imported can make local search within the prior knowledge, a new strategy is proposed in iteration strategy. The algorithm is validated for two test systems consisting of 6 and 15 generators. Compared with other methods in this literature, the experimental result demonstrates the high convergency and effectiveness of proposed algorithm.

In order to improve the problem that Artificial Bee Colony (ABC) is good at exploring but lack of exploitation, two new solution search strategies named PSO-DE-PABC and PSO-DE-GABC are proposed based on Particle Swarm Optimization (PSO) and Differential Evolution (DE). PSO-DE-PABC generates new candidate position around the random particle to improve divergence. PSO-DE-GABC generates new candidate position around the global best solution to accelerate the convergence, and differential vectors are also used to increase the divergence. Besides, Dimension Factor (DF) is introduced to control the search rate of the algorithms. A new scout strategy considering current swarm state is used to replace the original random scout strategy to enhance the local search ability. Comparison with basic ABC, GABC (Gbest-guided ABC) and ABC/best algorithm is given on 10 groups of standard benchmark function. The results show that PSO-DE-GABC and PSO-DE-PABC have better convergence rate and accuracy.

数据仓库与数据挖掘是大数据时代产生的一门新兴交叉的课程.针对该课程的特点,将CDIO工程教学理念融合到教学过程,重新设置了教学目标与大纲、调整了教学内容、改进了教学方法,总结了数据挖掘课程教学实践的一般流程并给出具体的实验教学设计方案.

Under the background of today's information age, microblog obtains a rapid development. With the news on the microblog updating, in order to avoid the users getting lost in the ocean of information, emotion analysis of the information becomes urgent and important. This paper based on the implementation of microblog emotion mining of Bayesian classifier and SVM classification algorithm, making comparison through the analysis of the experimental results in processing speed and accuracy, has a reference value.

在对外部事件影响下的网络舆情失控风险进行报警的过程中,各个因素相互影响并且各因素独立存在,在外部因素影响下复杂性更高,无法进行准确分类并设定波动较小的权值,导致传统的网络舆情失控风险报警方法,在利用决策树分类算法对舆论数据进行分类时候,不能客观、准确的对失控风险进行量化计算,无法实现失控风险的有效报警。提出基于外部事件分析的网络舆情失控风险报警模型,确定可能引起网络舆情失控的外部事件,通过被报警网络舆情的实际情况,对失控风险后果的属性类型进行确定,给出失控风险后果属性与其权重,对每种外部事件引起失控的可能性以及可能产生的后果值进行分析,获取所有外部事件影响网络舆情失控的相对严重程度,对其进行排序,报警实现外部事件影响下网络舆情失控风险的准确报警。仿真结果表明,所提方法能够准确的实现失控风险报警,更适合应用于实际网络失控风险报警中。

This paper studies evolutionary mechanism and information supervision of public opinions in Internet emergencies. The netizens ' behaviors are characterized by observation, imitation and learning, which well fits the hypotheses of bounded rationality in evolutionary games. In this paper, we define"netizen acceptance degree" in cases of some social hot topics as the payoff in factor games, and build the evolutionary equation of public opinions, which is the infectious diseases diffusion model with changing population size. After that, we ap-ply the theoretical model to the evolutionary course of public opinions in a real Internet emergency case, i. e. , the"2. 26 Self-sacrifice E-vent" in China. The effects of Internet information supervision and control measures for social affairs on the evolutionary course of public opinions are discussed based on results of numerical simulation.

This paper proposed a new method to train feedforward neural networks(FNNs) parameters based on the iterative chaotic map with infinite collapses particle swarm optimization(ICMICPSO) algorithm. This algorithm made full use of the information of BP’s error back propagation and gradient. It used ICMICPS as the global optimizer to adjust the neural networks’ weights and thresholds, when network parameters converge around global optimum. And it used gradient information as a local optimizer to accelerate the modification at a local scale. Compared with other algorithms, results show that the performance of the ICMICPSO-BPNN method is superior to the contrast methods in training and generalization ability.