Neural Network for Machine Learning-coursera-多伦多大学-无字幕
Neural Network for Machine Learning
https://www.coursera.org/learn/neural-networks 更多课件和练习，建议注册Hinton老师的课。
了解人工神经网络在语音识别和物体识别、图像分割、建模语言和人类运动等过程中的应用，及其在机器学习中发挥的作用。我们将重点介绍基本算法和使其充分发挥作用的实用技巧。
字幕有时会挡住字，所以没压（我才不说是嫌麻烦呢，哼）。
视频选集
（1/78）
001_Why do we need machine learning - [13 min]
13:15
002_What are neural networks - [8 min]
08:30
003_Some simple models of neurons [8 min]
08:24
004_A simple example of learning [6 min]
05:39
005_Three types of learning [8 min]
07:38
006_Types of neural network architectures [7 min]
07:29
007_Perceptrons - The first generation of neural networks [8 min]
08:17
008_A geometrical view of perceptrons [6 min]
06:25
009_Why the learning works [5 min]
05:10
010_What perceptrons can't do [15 min]
14:35
011_Learning the weights of a linear neuron [12 min]
11:55
012_The error surface for a linear neuron [5 min]
05:03
013_Learning the weights of a logistic output neuron [4 min]
03:57
014_The backpropagation algorithm [12 min]
11:51
015_Using the derivatives computed by backpropagation [10 min]
09:50
016_Learning to predict the next word [13 min]
12:34
017_A brief diversion into cognitive science [4 min]
04:27
018_Another diversion - The softmax output function [7 min]
07:21
019_Neuro-probabilistic language models [8 min]
07:52
020_Ways to deal with the large number of possible outputs [15 min]
12:17
021_Why object recognition is difficult [5 min]
04:40
022_Achieving viewpoint invariance [6 min]
05:58
023_Convolutional nets for digit recognition [16 min]
16:02
024_Convolutional nets for object recognition [17min]
17:44
025_Overview of mini-batch gradient descent
08:23
026_A bag of tricks for mini-batch gradient descent
13:16
027_The momentum method
08:43
028_Adaptive learning rates for each connection
05:44
029_Rmsprop - Divide the gradient by a running average of its recent magnitude
11:39
030_Modeling sequences - A brief overview
17:24
031_Training RNNs with back propagation
06:23
032_A toy example of training an RNN
06:15
033_Why it is difficult to train an RNN
07:44
034_Long-term Short-term-memory
09:16
035_A brief overview of Hessian Free optimization
14:25
036_Modeling character strings with multiplicative connections [14 mins]
14:36
037_Learning to predict the next character using HF [12 mins]
12:24
038_Echo State Networks [9 min]
09:37
039_Overview of ways to improve generalization [12 min]
11:45
040_Limiting the size of the weights [6 min]
06:23
041_Using noise as a regularizer [7 min]
07:32
042_Introduction to the full Bayesian approach [12 min]
10:50
043_The Bayesian interpretation of weight decay [11 min]
10:53
044_MacKay's quick and dirty method of setting weight costs [4 min]
03:32
045_Why it helps to combine models [13 min]
13:11
046_Mixtures of Experts [13 min]
13:16
047_The idea of full Bayesian learning [7 min]
07:28
048_Making full Bayesian learning practical [7 min]
06:44
049_Dropout [9 min]
08:35
050_Hopfield Nets [13 min]
13:02
051_Dealing with spurious minima [11 min]
11:02
052_Hopfield nets with hidden units [10 min]
09:40
053_Using stochastic units to improv search [11 min]
10:25
054_How a Boltzmann machine models data [12 min]
11:45
055_Boltzmann machine learning [12 min]
12:16
056_OPTIONAL VIDEO - More efficient ways to get the statistics [15 mins]
14:49
057_Restricted Boltzmann Machines [11 min]
10:54
058_An example of RBM learning [7 mins]
07:15
059_RBMs for collaborative filtering [8 mins]
08:17
060_The ups and downs of back propagation [10 min]
09:54
061_Belief Nets [13 min]
12:36
062_Learning sigmoid belief nets [12 min]
11:26
063_The wake-sleep algorithm [13 min]
13:15
064_Learning layers of features by stacking RBMs [17 min]
17:35
065_Discriminative learning for DBNs [9 mins]
09:41
066_What happens during discriminative fine-tuning - [8 mins]
08:39
067_Modeling real-valued data with an RBM [10 mins]
09:56
068_OPTIONAL VIDEO - RBMs are infinite sigmoid belief nets [17 mins]
17:12
069_From PCA to autoencoders [5 mins]
07:58
070_Deep auto encoders [4 mins]
04:11
071_Deep auto encoders for document retrieval [8 mins]
08:19
072_Semantic Hashing [9 mins]
08:50
073_Learning binary codes for image retrieval [9 mins]
09:38
074_Shallow autoencoders for pre-training [7 mins]
07:02
075_OPTIONAL - Learning a joint model of images and captions [10 min]
09:05
076_OPTIONAL - Hierarchical Coordinate Frames [10 mins]
09:41
077_OPTIONAL - Bayesian optimization of hyper-parameters [13 min]
13:30
078_OPTIONAL - The fog of progress [3 min]
02:25
顶部