一种神经网络开发程序集

支持向量机(3)svm_light-Support Vector Machines(3):svm_light

本人编写的incremental 随机神经元网络算法，该算法最大的特点是可以保证approximation特性，而且速度快效果不错，可以作为学术上的比较和分析。目前只适合benchmark的regression问题。 具体效果可参考 G.-B. Huang, L. Chen and C.-K. Siew, “Universal Approximation Using Incremental Constructive Feedforward Networks with Random Hidden Nodes”, IEEE Transactions on Neural Networks, vol. 17, no. 4, pp. 879-892, 2006. -I prepared by incremental random neural network algorithm, which is characterized by the largest approximation properties can be guaranteed, and fast good results can be used as an academic comparison and analysis. The current benchmark is only suitable for the regression problem. Specific effects may refer G.-B. Huang, L. Chen and C.-K. Siew,

室内移动机器人定位与导航关键技术研究,新思路-Indoor mobile robot localization and navigation key technology research and new ideas

此为疾病诊断专家系统，用这种结构建立的知识库可以很方便地实现上面所提到的各种功能。这种结 构的最独特之处就是依靠相同的关键（疾病名或其代码）来实现一种疾病与其症候之 间的相互联系。疾病名与症候分别用二叉树进行排序。排序后可由疾病名而查到与该 疾病名有关的所有详细情况，也可由一种症候而查到哪些疾病有这种症候，以及该症 候可能引起各种疾病的概率。-This is the disease diagnosis expert system, using this structure to establish the knowledge base can easily realize the above mentioned features. The structure of the most unique is that relying on the same key (the disease or its code) to achieve a disease and its symptoms between. Name and symptoms of disease were treated with binary tree sort. Sort by the disease after the name of the disease were found with all the details of which can also be a symptom of the diseases that have found such a symptom, as well as the symptoms may lead to the probability of various diseases.

此编码是一个数据挖掘的决策树各种算法。以作为入门提示-this code is a decision tree data mining algorithms. As a portal to suggest

这是一本进化算法的电子书，很适合人工智能方面的人学习-This is an evolutionary algorithm e-book, it is suitable for artificial intelligence to learn

这是一个有关于模式识别中的isodata算法的matlab源码。 大家可以看看。有详细注释-on the isodata pattern recognition algorithm Matlab source. We can look at. Detailed Notes

人体运动检测与运动跟踪的源代码，用到了最新的算法，对这方面研究的人员很有帮助-Human Motion Detection and Motion Tracking source code, use the latest algorithms, this study helpful staff

L-M算法。除了动量法(基于梯度下降的训练算法)外，学习率自适应调整策略是BP算法改进的另一种途径，它利用Levenberg-Marquardt优化方法，从而使得学习时间更短。其缺点是，对于复杂的问题，该方法需要很大的存储空间。 -L-M algorithm. In addition to momentum (based on the gradient descent algorithm for training), learning rate adaptive strategy is to improve the algorithm BP Another approach, which uses Levenberg-Marquardt optimization method, which makes learning time is even shorter. Its weakness is that the complex problems that the method requires a lot of storage space.

标准PSO（粒子群）算法，matlab版-standard PSO (PSO) algorithm, Matlab version