Python数据处理.pdf+代码

拟合线性函数y=mx+b，根据偏微分求导，梯度下降就是为了得到合适的线性参数，通过求导不断更新参数，来逐渐得到最接近的参数。先将系数随便设置一个数，将x带入，根据实际y值，调整系数。损失函数就是给定系数（一般初始值为0），得到预测值，将所有预测值减去真实值，就是损失函数。如果预测的很准，损失函数最小（为0）。

图卷积神经网络，图神经网络的一种，通过提取特征来训练神经网络，实现网络嵌入(Graph Convolution Neural Network, a kind of Graph Neural Network, trains the Neural Network by extracting features to realize network embedding.)

最近刚刚意识到python的重要性，学习中遇到并总结的经典知识记录在案(Recently, just realizing the importance of Python, the classic knowledge encountered and summarized in the study is recorded)

说明：  人工智能课程作业，TensorFlow中使用CNN实现手写体数字识别，基于CNN实现手写体数字识别并对比MLP分析。文章对代码原理结构有较为详尽的分析和解释，结尾处附有程序完整代码，可在python中直接运行。(Artificial intelligence course assignments, TensorFlow uses CNN to realize handwritten numeral recognition, and CNN to realize handwritten numeral recognition and MLP analysis. This article has a more detailed analysis and explanation of the code principle and structure. At the end of the article, there is a complete program code, which can run directly in python.)

【实例简介】 通过加载图像，自己输入水印文字和图像水印位置可以自己生成水印图片

软件无线电 MPSK调制python代码，还包括Gnuradio GRC文件(Software radio MPSK modulation code)

Excel 转为 json格式， python3 代码(Excel converted to json format, code written in python3)

说明：  层次分析法，简称AHP，是指将与决策总是有关的元素分解成目标、准则、方案等层次，在此基础之上进行定性和定量分析的决策方法。(It refers to the decision-making method that divides the elements that are always related to decision-making into objectives, criteria, schemes, etc. on the basis of which qualitative and quantitative analysis can be carried out)

图像处理(scipy.misc.bytescale.py)

说明：  这是一个Python基础的练习Demo,提供以后进行学习查看和记忆(this is a Python basic traning Demo,Provide for future learning, viewing and memory)