排序算法

机器学习的一系列算法的python实现，带实例(A series of machine learning algorithms Python implementation, with examples)

说明：  PyTorch是一个基于Torch的Python开源机器学习库，用于自然语言处理等应用程序，中文版官方教程是经典的PYTHON入门到深入教程(Pytorch is an open-source machine learning library for Python based on torch, which is used for natural language processing and other applications. The official Chinese version of the tutorial is a classic introduction to Python to in-depth tutorials)

说明：  使用tensorflow构建简单的神经网络，实现Mnist任务(We used tensorflow framework to build a simply neural network,and finished the task of Mnist classification.)

利用Python面向对象知识实现飞机大战游戏(Application of Aircraft Battle Game with Python)

构建了无向加权网，并对网络的度分布、集聚系数、最短路径、度匹配度、度中心性等重要参数进行了计算，在此基础上构建了社区挖掘算法用于网络的社区发现，并采用链路预测算法构建推荐系统(An undirected weighted network is constructed, and some important parameters, such as degree distribution, aggregation coefficient, shortest path, degree matching degree and degree centrality, are calculated. On this basis, a community discovery algorithm is built for the network, and a link prediction algorithm is used to construct a recommendation system.)

说明：  图片拼接 实现了多个图片的无缝拼接技术，完美无瑕(Picture stitching achieves seamless stitching technology of multiple pictures, perfect and flawless)

我们绘制图形主要用到两个库，matplotlib.pyplot和numpy。在编码过程中，这两个库的使用频率较高，而这两个库的名字较长。这难免会给我们带来不便。所以我们一般给其设置别名， 大大减少重复性工作量。 numpy 是 Python 用于数学运算的库，它是在安装 matplotlib 时候顺带安装的。pyplot 是 matplotlib 一个子模块，主要为底层的面向对象的绘图库提供状态机界面。状态机隐式地自动创建数字和坐标轴以实现所需的绘图。 matplotlib 中的所有内容都按照层次结果进行组织。顶层就是由 pyplot 提供的 matplotlib “状态机环境”。基于这个状态机环境，我们就可以创建图形。 作者：猴哥Yuri 链接：https://www.jianshu.com/p/78ba36dddad8 来源：简书 著作权归作者所有。商业转载请联系作者获得授权，非商业转载请注明出处。

实现了图像与视频的信息隐藏，输入图片和视频可进行效果演示(The information hiding of image and video is realized)

基于wordwind相同的遥感瓦片切分规则生成的瓦片数据，通过该程序可以还原成原有的tiff格式

说明：  用N来表示训练用例（样本）的个数，M表示特征数目。 输入特征数目m，用于确定决策树上一个节点的决策结果；其中m应远小于M。 从N个训练用例（样本）中以有放回抽样的方式，取样N次，形成一个训练集（即bootstrap取样），并用未抽到的用例（样本）作预测，评估其误差。 对于每一个节点，随机选择m个特征，决策树上每个节点的决定都是基于这些特征确定的。根据这m个特征，计算其最佳的分裂方式。 每棵树都会完整成长而不会剪枝，这有可能在建完一棵正常树状分类器后会被采用）。(N is used to represent the number of training cases (samples), and M is used to represent the number of features. The number of input features m is used to determine the decision result of a node in the decision tree, where m should be far less than m. From N training cases (samples), n times are sampled in the way of put back sampling to form a training set (i.e. bootstrap sampling), and the unselected cases (samples) are used to predict and evaluate the error. For each node, m features are randomly selected, and the decision of each node in the decision tree is determined based on these features. According to these m characteristics, the best splitting mode is calculated. Each tree will grow completely without pruning, which may be adopted after building a normal tree classifier).)