深度学习之TensorFlow：入门、原理与进阶实战

flume+kafka+storm整合，1.一个好的项目架构应该具备什么特点？ 2.本项目架构是如何保证数据准确性的？ 3.什么是Kafka？ 4.flume+kafka如何整合？ 5.使用什么脚本可以查看flume有没有往Kafka传输数据

低秩表达程序，MATLAB. Robust Subspace Segmentation by Low-Rank Representation(Low-Rank Representation program, MATLAB Robust Subspace Segmentation by Low-Rank Representation)

用R语言做dcc-garch 模型的全过程代码(Using the R language to do the dcc-garch model)

说明：  python PCA算法原理推导，及源代码实现，并基于PCA算法实现图片压缩，附pdf文件说明(Python PCA algorithm principle derivation, and source code implementation, and based on PCA algorithm to achieve image compression, with PDF file description)

用Python实现各种图算法,包括深度优先，广度优先，Dijkstral，Prim，WelshPowell_Coloring等算法，

说明：  深入理解Spark核心技术书籍，书籍对spark进行了深入讲解，并对spark源码进行了剖析(In-depth understanding of Spark core technology books, books on Spark in-depth explanation, and analysis of spark source code)

创建基于MySQL、Redis和MongoDB数据库开发命令行新闻管理系统。设定管理员和新闻编辑两种身份，实现发表新闻、编辑新闻、删除新闻、增加用户和删除用户等功能。

说明：  从数据库获取车辆在一段时间内的所有行驶记录的相关数据，确定所需数据为GPS经纬度坐标和驾驶时长等，QB模型采用MDF的思想,其基本思想为:通过平均直接翻转距离函数定义两条轨迹之间的距离，两条轨迹需要具有相同的经纬度点数，具有相同点数的轨迹最大的优点是对轨迹距离成对计算，且相同轨迹之间具有更高的分辨率，对于轨迹聚类的结果有一定的优化。(Retrieved from the database cars all over a period of time, record the related data, determine the required data for the GPS latitude and longitude coordinates, and the driving time, QB model by adopting the idea of MDF, its basic idea is: flip directly by the average distance function definition of the distance between two trajectories, two tracks will have the same latitude and longitude points, and has the biggest advantages of the same points of trajectory track distance calculation in pairs, and has higher resolution, between the same trajectory for trajectory clustering results have certain optimization.)

计算七类窗函数并给出归一化对数幅频曲线,同时也是利用窗函数法设计FIR滤波器的程序MDEFIR1所调用的子程序MWINDO(Seven kinds of window functions are calculated and normalized logarithmic amplitude-frequency curves are given. At the same time, it is also the subroutine MWINDO called by the program MDEFIR1 of designing FIR filter by using window function method.)

本程序主要是针对超大整型数据的基础加、减、乘运算，运算位数可达10000位，同时也支持使用者进行修改扩展至更大位数