资源列表
《神经网络与深度学习》讲义-邱锡鹏
- 《神经网络与深度学习》讲义-邱锡鹏 一个人在不接触对方的情况下，通过一种特殊的方式，和对方进行 一系列的问答。如果在相当长时间内，他无法根据这些问题判断对方是 人还是计算机，那么就可以认为这个计算机是智能的。(Lectures on neural networks and deep learning - Qiu Xipeng A person who has no contact with each other has a series of questions and answers wi
《神经网络与深度学习》Michael Nielsen
- 这本书最初是我学习 Neural Networks and Deep Learning 时做的中?笔记，因为原书中有很 A A 多数学公式，所以我? LT E X 来编写和排版，并将所有 LT E X 源码放置在 GitHub。其中部分内容 取? Xiaohu Zhu 已经完成的翻译来避免重复的?作。(This book was originally Neural Networks and Deep Learning I learned to do in notes, because there
Python-ELM-master
- 极限学习机的 python实现，可用于回归，多分类问题等。(The python coad of the extreme learning machine which can be used for regression, multiple classification problems, etc.)
tensorflow-cnn
- 基于TensorFlow的mnist数据集识别，使用CNN的方法，采用梯度下降学习(MNIST data set recognition based on TensorFlow, using CNN method, using gradient descent learning)
chap7
- 通过BP神经网络训练实现分类，将三维的输入转化为固定的二维输出(Through the BP neural network training to achieve classification, the three-dimensional input into a fixed two-dimensional output)
jhabb
- Normalized data model, modal vibration, MinkowskiMethod algorithm, Build a framework OFDM communication system.
DFA应用
- ACM资料，来自北京大学的暑假培训讲课资料，希望能对大家有所帮助(ACM information, from the Peking University summer training, lectures, information, we hope to help)
mscnn-master
- MSCNN多尺度rcnn的目标检测代码，可以实现多尺度的目标检测，具有多尺度的感受野(MSCNN multi-scale RCNN target detection code)
BP
- BP神经网络经过训练与学习，逼近非线性函数(BP neural networks are trained and studied to approximate nonlinear functions)
apriori
- 数据挖掘Aprior的简易c++实现，通过string来表示数据集，没有采用STL set集合操作，因此实现起来简单一些，判断是否包含用到了动态规划算法中的最长公共子序列，思想是暴力求解，没有用到很高端的思想和算法。(Simple c++ data mining Aprior, said data set by the string, without the use of STL set collection operation, thus the realization is simple,
BPPIDspeed
- 神经网络BP PID 控制双闭环直流电机(Neural network, BP, PID control, double closed loop, DC motor)
CNN Matlab代码
- 利用大量图像数据对卷积神经网络算法进行训练，通过卷积、池化、下采样以及全连接层训练后的卷积神经网络在图像识别精度越来越高。(By using a large number of image data to train the convolutional neural network algorithm, the accuracy of the image recognition is higher and higher by convolution, pooling, down sampling