1. 池化和采样 Pooling Max/Avg pooling upsample nearest bilinear Relu 2. cifar100 13 layers Code import tensorflow as tf from tensorflow.keras import layers, optimizers, metrics, dat...

1. 如何通俗易懂地解释卷积？ 转载自知乎，马同学 2. 2D Convolution 3. Padding & Stride 4. Channels 5. layers.Conv2D 6. 需要全套课程视频+PPT+代码资源可以私聊我 方式1：优快云私信我! 方式2：QQ邮箱:594042358@qq.com或者直接加我QQ: 594042358! ...

1. Keras.Metrics （度量指标） 1.1. Build a meter 1.2. Update data 1.3. Get Average data 1.4. Clear buffer 1.5. Code import tensorflow as tf from tensorflow.keras import datasets, layers, optimizers, Se...

1. 梯度下降（Gradient Descent） 1.1. What’s Gradient 1.2. What does it mean 1.3. How to Search 1.4. AutoGrad GradientTape Persistent GradientTape 2nd-order import tensorflow as tf w = tf.Var...

1. 均方误差MSE 2. 交叉熵损失Cross Entropy Loss 2.1. 信息熵/信息增益 2.2. Cross Entropy 2.3. Classification 2.4. Why not MSE？ 2.5. svm损失函数Hinge Loss 3. logits→CrossEntropy 4. Numerical Stability import tenso...

1. Outline 1.1. Matmul （矩阵形式） ???????????? = ????(????@???? + ????) ???????????? = ????????????????(????@???? + ????) X@W+b 1.2. Neural Network （神经网络结构） lecun在86年89年在手写数字论文中用到的就是这个网络， 当时叫做神经网络。 3~5 layers 1.3. Deep Learning （深度学习） n 约等于 ...

1. tensorflow.keras.datasets 1.1. 常用数据集 boston housing （波士顿的房价） Boston housing price regression dataset. mnist/fashion mnist （手写数字） MNIST/Fashion-MNIST dataset. cifar10/100 （小型的图片识别数据集） smal...

123

1. 前向传播案例 1.1. 代码实现 import tensorflow as tf from tensorflow import keras from tensorflow.keras import datasets # 加载数据集 # x: [60k, 28, 28] # y: [60k] 0~9 (x, y), _ = datasets.mnist.load_data() # x:[...

1. Outline ±*/ **, pow, square sqrt //, % exp, log @, matmul linear layer 2. Operation type element-wise ±*/ matrix-wise @, matmul dim-wise reduce_mean/max/min/sum 2.1. ±*/%// 2.2. tf.ma...

1. Broadcasting 1.1. Outline expand without copying data VS tf.tile tf.broadcast_to 1.2. Key idea Insert 1 dim ahead if needed Expand dims with size 1 to same size Feature maps: [4, 32, 32, 3] B...

1. Outline shape, ndim reshape expand_dims/squeeze transpose 2. Reshape 2.1. Reshape is flexible 2.2. Reshape could lead to potential bugs! 3. tf.transpose 4. Squeeze VS Expand_dims 4.1. Expand ...

1. Indexing 1.1. Basic indexing 1.2. Same With Numpy 1.2.1. Numpy-style indexing 1.2.2. start:end 1.2.3. Indexing by : 1.2.4. Indexing by :: 1.2.5. ::-1 1.2.6. …

1. 数据类型 1.1. 数据容器 list np.array tf.Tensor 1.2. What’s Tensor scalar: 1.1 vector: [1.1],[1.1, 2.2, … ] matrix: [[1.1, 2.2],[3.3, 4.4],[5.5, 6.6]] tensor: ???????????????? > 2 (在神经网络中， tensor的范围相当广， 这里可以理解为狭义...

1. 手写数字识别简述 1.1. MNIST 7000 images per category train/test splitting: 60k vs 10k 1.2. Image [28,28,1] →[784] 1.3. Input and Output 1.4. Regression VS Classification ???? = ???? ∗ ???? + ???? ???? ∈ ???????? ????????????...

1. 线性回归 1.1. Continuous Prediction 1.2. Linear Equation 1.3. With Noise? 1.4. For Example 1.5. Find ????′, ????′ 1.6. Gradient Descent 2. Numpy实现线性回归 2.1 Find ????′, ????′ 2.2. Solution Compute Loss Comp...

1. 发展历史 1.1. 前世今生 2015.9发布0.1版本 2017.2发布1.0版本 2019春发布2.0版本 1.2. 同一时期人工智能其他库的发展 1.2.1. 2015年 Scikit-learn Machine learning, No GPU Caffe 2013, 第一个面向深度学习的框架 No auto-grad, C++ Keras wrapper T...