采用DNA进化算法优化设计TS模糊控制器

The DNA evolutionary algorithm(DNA-EA)and the DNA genetic algorithm(DNA-GA)based on a new DNA encoding method are propsed based on the structure and the genetic mechanism of biological DNA. The DNA-EA and the DNA-GA are applied into the optimal design of TS fuzzy control system. The simulation results show the effectiveness of the two DNA algorithms, excellent self-learning capability. However, the DNA-EA is superior to the DNA-GA in the simulation performance.     

Optimal decision rules acquisition in incomplete and Fuzzy Decision Information System

The Incomplete and Fuzzy Decision Information System （IFDIS） with both missing data and fuzzy decisions is rather extensively applications, which is the generalizations of complete information system, incomplete information system and fuzzy decision information system. In this paper, on the base of notion of the tolerance relation and the concept of IFDIS, the rough set model of IFDIS based on tolerance relation is proposed and the properties of the models are discussed, and then the definitions of（upper- or lower-） consistent reduction are suggested and knowledge reduction based on discernibility matrix is presented. The optimal fuzzy decision rules and its acquisition algorithm are proposed. Finally we provide an example to illustrate the validity of the algorithm.     

An Adaptive Learning Method for the Generation of Fuzzy Inference System from Data

Designing a fuzzy inference system （FIS） from data can be divided into two main phases： structure identification and parameter optimization. First, starting from a simple initial topology, the membership functions and system rules are defined as specific structures. Second, to speed up the convergence of the learning algorithm and lighten the oscillation, an improved descent method for FIS generation is developed. Furthermore, the convergence and the oscillation of the algorithm are systematically analyzed. Third, using the information obtained from the previous phase, it can be decided in which region of the input space the density of fuzzy rules should be enhanced and for which variable the number of fuzzy sets that used to partition the domain must be increased. Consequently, this produces a new and more appropriate structure. Finally, the proposed method is applied to the problem of nonlinear function approximation.     

Interactive image enhancement by fuzzy relaxation

In this paper,an interactive image enhancement(IIE)technique based on fuzzy relaxation is presented,which allows the user to select different intensity levels for enhancement and intermit the enhancement process according to his/her preference in applications.First,based on an analysis of the convergence of a fuzzy relaxation algorithm for image contrast enhancement,an improved version of this algorithm,which is called FuzzIIE Method 1,is suggested by deriving a relationship between the convergence regions and the parameters in the transformations defined in the algorithm.Then a method called FuzzIIE Method 2 is introduced by using a different fuzzy relaxation function,in which there is no need to re-select the parameter values for interactive image enhancement. Experimental results are presented demonstrating the enhancement capabilities of the proposed methods under different conditions.     

A novel multiresolution fuzzy segmentation method on MR image

Multiresolution-based magnetic resonance (MR) image segmentation has attracted attention for its ability to capture rich information across scales compared with the conventional segmentation methods. In this paper, a new scale-space-based segmentation model is presented,where both the intra-scale and inter-scale properties are considered and formulated as two fuzzy energy functions. Meanwhile, a control parameter is introduced to adjust the contribution of the similarity character across scales and the clustering character within the scale. By minimiT.ing thecombined inter/intra energy function, the multiresolution fuzzy segmentation algorithm is derived.Then the coarse to fine leading segmentation is performed automatically and iteratively on a set of multiresolution images. The validity of the proposed algorithm is demonstrated by the test image and pathological MR images. Experiments show that by this approach the segmentation results,especially in the tumor area delineation, are more precise than those of the conventional fuzzy segmentation methods.     

一类基于数据的解释性模糊建模方法的研究

An approach to identify interpretable fuzzy models from data is proposed. Interpretability, which is one of the most important features of fuzzy models, is analyzed first. The number of fuzzy rules is determined by fuzzy cluster validity indices. A modified fuzzy clustering algorithm,combined with the least square method, is used to identify the initial fuzzy model. An orthogonal least square algorithm and a method of merging similar fuzzy sets are then used to remove the redundancy of the fuzzy model and improve its interpretability. Next, in order to attain high accuracy, while preserving interpretability, a constrained Levenberg-Marquardt method is utilized to optimize the precision of the fuzzy model. Finally, the proposed approach is applied to a PH neutralization process, and the results show its validity.     

Universal triple I fuzzy reasoning algorithm of function model based on quotient space

Aiming at the deficiencies of analysis capacity from different levels and fuzzy treating method in product function modeling of conceptual design,the theory of quotient space and universal triple I fuzzy reasoning method are introduced,and then the function modeling algorithm based on the universal triple I fuzzy reasoning method is proposed.Firstly,the product function granular model based on the quotient space theory is built,with its function granular representation and computing rules defined at the same time.Secondly,in order to quickly achieve function granular model from function requirement,the function modeling method based on universal triple I fuzzy reasoning is put forward.Within the fuzzy reasoning of universal triple I method,the small-distance-activating method is proposed as the kernel of fuzzy reasoning;how to change function requirements to fuzzy ones,fuzzy computing methods,and strategy of fuzzy reasoning are respectively investigated as well;the function modeling algorithm based on the universal triple I fuzzy reasoning method is achieved.Lastly,the validity of the function granular model and function modeling algorithm is validated.Through our method,the reasonable function granular model can be quickly achieved from function requirements,and the fuzzy character of conceptual design can be well handled,which greatly improves conceptual design.     

Adaptive sliding mode-like fuzzy logic control for nonlinear systems

Sliding mode-like fuzzy logic control （SMFC） algorithm for nonlinear systems is presented in this paper. Firstly dead zone parameters of sliding mode control （SMC） are selftuned by proper adaptive laws and then combined into fuzzy logic system （FLS） to compose the opportune fuzzy logic control （FLC）, which is equivalent to the predesigned SMC controller with self-tuning parameters. Robustness and invariance to the uncertainties of the closed-loop systems are improved and chattering of the SMC is eliminated. Finally simulation results of numerical examples show that the proposed control algorithm is efficient and feasible.     

Relaxed delay-dependent stabilization criteria for discrete-time fuzzy systems based on a switching fuzzy model and piecewise Lyapunov functions

This paper presents delay-dependent stability analysis and controller synthesis methods for discrete-time Takagi-Sugeno （T-S） fuzzy systems with time delays. The T-S fuzzy system is transformed to an equivalent switching fuzzy system. Consequently, the delay-dependent stabilization criteria are derived for the switching fuzzy system based on the piecewise Lyapunov function. The proposed conditions are given in terms of linear matrix inequalities （LMIs）. The interactions among the fuzzy subsystems are considered in each subregion, and accordingly the proposed conditions are less conservative than the previous results. Since only a set of LMIs is involved, the controller design is quite simple and numerically tractable. Finally, a design example is given to show the validity of the proposed method.     

Double-layer bayesian classifier ensembles based on frequent itemsets

Numerous models have been proposed to reduce the classification error of Na¨ ve Bayes by weakening its attribute independence assumption and some have demonstrated remarkable error performance. Considering that ensemble learning is an effective method of reducing the classification error of the classifier, this paper proposes a double-layer Bayesian classifier ensembles (DLBCE) algorithm based on frequent itemsets. DLBCE constructs a double-layer Bayesian classifier (DLBC) for each frequent itemset the new instance contained and finally ensembles all the classifiers by assigning different weight to different classifier according to the conditional mutual information. The experimental results show that the proposed algorithm outperforms other outstanding algorithms.     

一种适用于多类不平衡数据集的模糊关联分类方法

提出一种适用于多类不平衡分布情形下的模糊关联分类方法,该方法以最小化AdaBoost.M1W集成学习迭代过程中训练样本的加权分类错误率和子分类器中模糊关联分类规则数目及规则中所含模糊项的数目为遗传优化目标,实现了AdaBoost.M1W和模糊关联分类建模过程的较好融合.通过5个多类不平衡UCI标准数据集和现有的针对不平衡分类问题的数据预处理方法实验对比结果,表明了所提出的方法能显著提高多类不平衡情形下的模糊关联分类模型的分类性能.     

基于混合策略的关联分类方法

关联分类中现有的显式学习方法无法解决small disjunction问题,而Lazy方法分类效率低。针对这两类方法存在的问题,提出了一种基于混合策略的关联分类方法。具体算法为:先判断待分类样本是否满足显式学习模式的分类器特征;然后把满足分类器特征的待分类样本用显式模式进行分类,把不满足分类器特征的待分类样本用Lazy模式来预测;最后结合两类方法的分类结果得到最终的分类结果。实验比较了该方法与传统的关联分类方法,结果表明,该方法在分类准确率和执行效率方面均达到了更好的效果。     

一种模糊分类器集成的方法

提出了一种基于模糊积分的模糊分类器集成的方法，该方法能在模糊分类器生成过程中，进一步减少主观因素的参与成份，使分类模器具有更好的稳定性和更高的分类识别率。给出了基于隶属度矩阵的模糊积分密度确定方法，介绍了基于模糊积分的分类器集成算法。用权威的数据集作为实验数据集，将提出方法与已有的分类器集成方法进行实验比较，评测了所提出方法的有效性。     

Support Vector的几何解释及其在联想记忆中的应用

利用闭凸集上的投影解释support vector的几何意义，利用支持超平面讨论线性分类器的设计问题，对线性可分情形，Support vector由一类数据集合闭凸包在另一类数据集合闭凸包上投影的非零系数向量组成，SVM所决定的超平面位于两投影点关于各自数据集合支持超平面的中间，作为应用，文中给出一种设计理想联想记忆前馈神经网络的方法，它是FP算法的一般化。     

一种基于模糊关联分类的遥感图像分类方法

遥感图像分类是遥感领域的研究热点之一.提出了一种基于自适应区间划分的模糊关联遥感图像分类方法(fuzzy associative remote sensing classification,FARSC).算法根据遥感图像分类的特点,利用模糊C均值聚类算法自适应地建立连续型属性模糊区间,使用新的剪枝策略对项集进行筛选从而避免生成无用规则,采用一种新的规则重要性度量方法对多模糊分类规则进行融合,从而有效地提高分类效率和精确度.在UCI数据和遥感图像上所作实验结果表明,算法具有较高的分类精度以及对样本数量变化的不敏感性,对于解决遥感图像分类问题,FARSC算法具有较高的实用性,是一种有效的遥感图像分类方法.     

一种大数据环境中分布式辅助关联分类算法

在很多现实的分类应用中,新数据的类标需要由领域专家最终确定,而分类器的分类结果仅起辅助作用.另外,随着大数据所隐含价值越发被人们重视,分类器的训练会从面向单一数据集逐渐过渡到面向分布式空间数据集,大数据环境下辅助分类也将成为未来分类应用的重要分支.然而,现有的分类研究缺乏对此类应用的关注.大数据环境中的辅助分类面临以下3个问题:1) 训练集是分布式大数据集;2) 在空间上,训练集所包含的各局部数据源的类别分布不尽相同;3) 在时间上,训练集是动态变化的,会发生类别迁移现象.在考虑以上问题的基础上,提出一种大数据环境中分布式辅助关联分类方法.该方法首先给出一种大数据环境中分布式关联分类器构建算法,在该算法中,通过横向加权考虑分类数据集在空间上的类别分布差异,并给出"前件空间支持度-相关系数"的度量框架,改进关联分类算法面对不平衡数据的性能缺陷;然后,给出一种基于适应因子的辅助关联分类器动态调整方法,能够在分类器应用过程中充分利用领域专家实时反馈的结果对分类器进行动态调整,以提升其面向动态数据集的分类性能,减缓分类器的退化和重新训练的频率.实验结果表明,该方法能够面向分布式数据集较快地训练出有较高分类准确率的关联分类器,并在数据集不断扩充变化时提升分类性能,是一种有效的大数据环境中辅助分类应用方法.     

基于可分度和支持度的模糊密度赋值融合识别算法

模糊积分理论可有效处理分类决策不确定性问题。当前模糊密度的确定方法未考虑各个分类器识别结果的可区分程度及各分类器对识别结果的支持程度,会丢失融合识别的相关信息。文中提出基于可分度和支持度的自适应模糊密度赋值融合识别算法。该算法根据各分类器对待识别样本的识别结果的可区分程度和支持程度对分类器的融合模糊密度进行自适应赋值,从而有效实现多分类器融合识别。将该算法应用于自然交互环境下的人脸表情识别和Cohn-Kanade表情识别。实验结果表明,该算法能有效提高总体表情识别率。     

基于线性判别分析的Choquet积分的符号模糊测度提取

在解决分类问题时,建立在Choquet积分上的分类器以其非线性和不可加性的特点,扮演着越来越重要的角色。由于Choquet积分中的符号模糊测度可以描述各特征对结果的影响,因此Choquet积分在解决数据分类及融合 问题方面具有显著的优势。但是,关于Choquet积分符号模糊测度值的求解,学术界一直缺乏有效的方法。目前最常用的方法是遗传算法,但是遗传算法在解决符号模糊测度值的优化问题时存在算法较为复杂、耗时较长等缺陷。由于符号模糊测度值在Choquet积分分类器中是决定性的重要参数,因此设计出一种有效的符号模糊测度提取方法十分必要。文中提出基于线性判别分析的Choquet积分符号模糊测度的提取方法,推导出在分类问题下Choquet积分的符号模糊测度值的解析式表达,其能够有效、快速地得出关键性参数。分别在人工数据集及基准实际数据集上进行测试与验证,实验结果表明所提方法能有效解决Choquet积分分类器中符号模糊测度的优化问题。     

模糊联想记忆网络的全局鲁棒性研究——基于爱因斯坦t-模

利用三角模的模糊联想记忆网络的性质以及模糊联想记忆网络的鲁棒性定义,对基于爱因斯坦t-模构建的模糊双向联想记忆网络的学习算法的全局鲁棒性进行了分析。从理论上证明了当训练模式的摄动为正向摄动时,该学习算法可以保持良好的鲁棒性,并用实验验证了该结论;而当摄动存在负向波动时该学习算法不满足全局鲁棒性。然后又进一步对训练模式集摄动最大摄动与输出模式集的最大摄动之间的关系进行研究,得出了训练模式集的最大摄动与输出模式集的最大摄动之间的关系曲线。