▍1. Python数据分析基础教程:NumPy学习指南.第2版
说明: python数据分析基础教程,为实现利用python进行数据分析提供基础入门介绍(The basic course of Python data analysis provides a basic introduction to the implementation of data analysis with Python)
说明: python数据分析基础教程,为实现利用python进行数据分析提供基础入门介绍(The basic course of Python data analysis provides a basic introduction to the implementation of data analysis with Python)
说明: 一维连续小波函数和信号领域深度学习的结合(The combination of one-dimensional continuous wavelet function and deep learning in signal field)
说明: 针对序列信号分类识别Classification and recognition of sequence signals(Classification and recognition of sequence signals)
说明: deep+learning深度学习人工智能资料(deep+learning Deep learning of AI materials)
说明: 机器学习是计算机科学与人工智能的重要分支领域. 本书作为该领域的入门教材,在内容上尽可能涵盖机器学习基础知识的各方面。 为了使尽可能多的读者通过本书对机器学习有所了解, 作者试图尽可能少地使用数学知识. 然而, 少量的概率、统计、代数、优化、逻辑知识似乎不可避免. 因此, 本书更适合大学三年级以上的理工科本科生和研究生, 以及具有类似背景的对机器学 习感兴趣的人士. 为方便读者, 本书附录给出了一些相关数学基础知识简介。(Machine learning is an important branch of computer science and artificial intelligence. As an introductory textbook in this field, this book covers all aspects of basic knowledge of machine learning as much as possible. In order to make as many readers as possible understand machine learning through this book, the author tries to use mathematical knowledge as little as possible. However, a small amount of probability, statistics, algebra, optimization, logic knowledge seems inevitable. Therefore, this book is more suitable for undergraduates and postgraduates of science and engineering above the third grade of University, And people who are interested in machine learning with similar background. For the convenience of readers, the appendix of this book gives a brief introduction of some basic mathematical knowledge.)
说明: 通过python,用卷积神经网络实现手写汉字数字的识别(Recognition of handwritten Chinese characters by convolutional neural network)
说明: 训练一个卷积神经网络,用fastai库(建在PyTorch上)将图像分类为纸板,玻璃,金属,纸张,塑料或垃圾。数据集下载地址如下。(Train a convolutional neural network to classify images into cardboard, glass, metal, paper, plastic or garbage using the fastai Library (built on pytorch). The data set download address is as follows.)
说明: 自定义的mininet-wifi的网络拓扑(Custom mininet-wifi network topology)
说明: 使用神经网络进行训练,对高光谱普图像进行分类(Using neural network to train and classify hyperspectral images)
说明: 对高光谱图像分类,利用光谱图像对高光谱图像进行处理(Classification of light spectrum image)
说明: 利用深度学习卷积神经网络,对手写数字1到9进行训练,使得训练网络识别数字的准确率达到比较高的水平(Deep learning convolutional neural network is used to train the hand to write Numbers from 1 to 9, so that the accuracy of the training network to recognize Numbers can reach a relatively high level)
说明: 卷积神经网络分类 调制信号识别 卷积神经网络(Convolutional Neural Networks, CNN)是一类包含卷积计算且具有深度结构的前馈神经网络(Feedforward Neural Networks),是深度学习(deep learning)的代表算法之一 [1-2] 。卷积神经网络具有表征学习(representation learning)能力,能够按其阶层结构对输入信息进行平移不变分类(shift-invariant classification),因此也被称为“平移不变人工神经网络(Shift-Invariant Artificial Neural Networks, SIANN)” [3] 。 对卷积神经网络的研究始于二十世纪80至90年代,时间延迟网络和LeNet-5是最早出现的卷积神经网络 [4] ;在二十一世纪后,随着深度学习理论的提出和数值计算设备的改进,卷积神经网络得到了快速发展,并被应用于计算机视觉、自然语言处理等领域 [2] 。 卷积神经网络仿造生物的视知觉(visual perception)机制构建,可以进行监督学习和非监督学习,其隐含层内的卷积核参数共享和层间连接的稀疏性使得卷积神经网络能够以较小的计算量对格点化(grid-like topology)特征,例如像素和音频进行学习、有稳定的效果且对数据没有额外的特征工程(feature engineering)要求 [1-2] 。(Convolutional neural network classification modulation signal recognition)
说明: 应用:森林应用单木分割python代码下载,算法下载,处理点云数据进行单木分割。(orest application single wood segmentation Python code download, algorithm download, processing point cloud data single wood segmentation.)
说明: 一网打尽神经序列模型之RNN及其变种LSTM、GRU(RNN and its variants, LSTM and Gru, in the model of one net out neural sequence)
说明: 采用MNE处理脑电数据,利用随机森林进行分类的睡眠分类实例(An example of sleep classification using MNE to process EEG data and random forest to classify)
说明: 采用一维卷积神经网络对癫痫脑电信号段进行分类(Classification of epileptic EEG signal segments by one dimensional convolutional neural network)
说明: 基于udp的网络聊天室,使用python语言,具有注册登录公频聊天和私聊等功能。(The Internet chat room based on UDP, using Python language, has the functions of registering and logging in public frequency chat and private chat.)
说明: 卷积神经网络与支持向量机结合的python代码(Python code combining convolutional neural network and support vector machine)
说明: 针对数据集较少导致人体动作识别率较低的情况,基于生成对抗网络搭建了一个人体动作识别网络,有效解决了以上问题(In view of the low recognition rate of human motion caused by the lack of data set, a human motion recognition network is built based on the generated countermeasure network, which effectively solves the above problems)