Hugo Larochelle 教授的神经网络课程
https://www.youtube.com/watch?v=SGZ6BttHMPw&list=PL6Xpj9I5qXYEcOhn7TqghAJ6NAPrNmUBH
这是Hugo Larochelle 教授的神经网络课程视频, 做得非常棒, 虽然是生肉, 但其实也很容易听懂, 课程本身也深入浅出, 慢慢看对学习非常有帮助.
视频选集
（1/92）
自动连播
01 01 Feedforward neural network - artificial neuron
07:50
01 02 Feedforward neural network - activation function
05:55
01 03 Feedforward neural network - capacity of single neuron
08:04
01 04 Feedforward neural network - multilayer neural network
13:10
01 05 Feedforward neural network - capacity of neural network
08:55
01 06 Feedforward neural network - biological inspiration
14:20
02 01 Training - empirical risk minimization
10:27
02 02 Training - loss function
04:48
02 03 Training - output layer gradient
12:02
02 04 Training - hidden layer gradient
15:14
02 05 Training - activation function derivative
04:36
02 06 Training - parameter gradient
06:25
02 07 Training - backpropagation
15:05
02 08 Training - regularization
13:14
02 09 Training - parameter initialization
06:09
02 10 Training - model selection
13:47
02 11 Training - optimization
23:39
03 01 Conditional random fields - motivation
05:18
03 02 Conditional random fields - linear chain CRF
09:57
03 03 Conditional random fields - context window
12:46
03 04 Conditional random fields - computing the partition function
24:33
03 05 Conditional random fields - computing marginals
09:07
03 06 Conditional random fields - performing classification
18:31
03 07 Conditional random fields - factors sufficient statistics and linear CRF
11:36
03 08 Conditional random fields - Markov network
11:36
03 09 Conditional random fields - factor graph
06:27
03 10 Conditional random fields - belief propagation
24:47
04 01 Training CRFs - loss function
05:44
04 02 Training CRFs - unary log-factor gradient
13:28
04 03 Training CRFs - pairwise log-factor gradient
05:53
04 04 Training CRFs - discriminative vs generative learning
06:43
04 05 Training CRFs - maximum-entropy Markov model
08:45
04 06 Training CRFs - hidden Markov model
04:16
04 07 Training CRFs - general conditional random field
06:29
04 08 Training CRFs - pseudolikelihood
05:10
05 01 Restricted Boltzmann machine - definition
12:16
05 02 Restricted Boltzmann machine - inference
18:31
05 03 Restricted Boltzmann machine - free energy
12:53
05 04 Restricted Boltzmann machine - contrastive divergence
13:33
05 05 Restricted Boltzmann machine - contrastive divergence parameter update
11:09
05 06 Restricted Boltzmann machine - persistent CD
07:35
05 07 Restricted Boltzmann machine - example
08:14
05 08 Restricted Boltzmann machine - extensions
09:18
06 01 Autoencoder - definition
06:14
06 02 Autoencoder - loss function
11:51
06 03 Autoencoder - example
02:53
06 04 Autoencoder - linear autoencoder
19:46
06 05 Autoencoder - undercomplete vs overcomplete hidden layer
05:35
06 06 Autoencoder - denoising autoencoder
14:15
06 07 Autoencoder - contractive autoencoder
12:07
07 01 Deep learning - motivation
15:11
07 02 Deep learning - difficulty of training
08:23
07 03 Deep learning - unsupervised pre-training
12:51
07 04 Deep learning - example
12:40
07 05 Deep learning - dropout
11:17
07 06 Deep learning - deep autoencoder
07:33
07 07 Deep learning - deep belief network
13:21
07 08 Deep learning - variational bound
14:02
07 09 Deep learning - DBN pre-training
19:59
08 01 Sparse coding - definition
12:04
08 02 Sparse coding - inference ISTA algorithm
12:35
08 03 Sparse coding - dictionary update with projected gradient descent
05:03
08 04 Sparse coding - dictionary update with block-coordinate descent
13:09
08 05 Sparse coding - dictionary learning algorithm
05:30
08 06 Sparse coding - online dictionary learning algorithm
09:04
08 07 Sparse coding - ZCA preprocessing
08:38
08 08 Sparse coding - feature extraction
10:42
08 09 Sparse coding - relationship with V1
05:45
09 01 Computer vision - motivation
05:24
09 02 Computer vision - local connectivity
04:18
09 03 Computer vision - parameter sharing
11:31
09 04 Computer vision - discrete convolution
15:26
09 05 Computer vision - pooling and subsampling
08:10
09 06 Computer vision - convolutional network
13:57
09 07 Computer vision - object recognition
07:59
09 08 Computer vision - example
14:19
09 09 Computer vision - data set expansion
07:30
09 10 Computer vision - convolutional RBM
10:45
10 01 Natural language processing - motivation
02:15
10 02 Natural language processing - preprocessing
09:45
10 03 Natural language processing - one-hot encoding
07:30
10 04 Natural language processing - word representations
10:29
10 05 Natural language processing - language modeling
09:22
10 06 Natural language processing - neural network language model
16:07
10 07 Natural language processing - hierarchical output layer
13:50
10 08 Natural language processing - word tagging
10:47
10 09 Natural language processing - convolutional network
16:43
10 10 Natural language processing - multitask learning
16:02
10 11 Natural language processing - recursive network
05:49
10 12 Natural language processing - merging representations
03:39
10 13 Natural language processing - tree inference
16:50
10 14 Natural language processing - recursive network training
13:28
顶部