0.25
0.5
0.75
1.25
1.5
1.75
2
Multilayer Neural Networks
Published on Sep 13, 201522738 Views
Related categories
Chapter list
Multilayer Neural Networks00:00
Success stories01:04
Part 101:35
Part 201:57
Part 302:14
Neural Information Processing02:24
The perceptron - 102:25
The perceptron - 202:28
The perceptron is a machine03:08
The perceptron - 303:37
Cybernetics (1948)04:09
How to design computers?04:57
Computing with symbols05:27
Computing with the brain05:54
McCulloch & Pitts (1943)06:52
Perceptrons (1968)07:45
Perceptrons revisited10:29
Neural Information Processing12:21
Quillian’s hierarchical propositional model (1968) - 112:40
Quillian’s hierarchical propositional model (1968) - 213:42
Connectionism - 114:45
Connectionism - 216:08
Training the network18:06
Propagation18:08
Back-Propagation18:33
Training algorithm (batch)19:27
Training algorithm (stochastic)20:15
Outputs20:44
Representations - 121:32
Representations - 222:00
ISA in representation space22:17
Dynamic reconfiguration - 122:48
Dynamic reconfiguration - 224:14
Network construction kit25:09
Linear brick25:14
Transfer function brick26:00
Transfer functions26:12
Square loss brick26:46
Loss bricks27:11
Sequential brick27:35
Benefits28:07
Torch code sample - 129:17
Torch code sample - 229:47
Convolutional networks (CNNs)31:08
Vision is fast32:59
Hubel & Wiesel (1962)33:55
The Neocognitron34:50
Local connections36:01
Convolution36:29
Multiple convolutions37:11
CNNs in the 1990s37:18
Convnets in the 1990s39:18
Pooling39:49
Contrast Normalization41:10
CNNs in the 2010s41:49
Torch code sample - 342:07
Convnets in the 2000s42:36
ImageNet 2012 competition42:52
ImageNet CNN - 143:56
ImageNet CNN - 244:30
Replicated CNNs44:44
Replicated CNNs at work45:38
CNNs for speech recognition - 146:13
CNNs for speech recognition - 248:11
CNNs for speech recognition - 348:48
Optimization basics49:02
Convex49:16
Non-convex50:12
Derivatives52:30
Line search - 153:58
Line search - 254:33
Line search - 354:47
Line search - 454:55
Line search - 554:56
Line search - 654:57
Line search - 755:34
Line search - 955:50
Line search - 1056:08
Parabola01:01:01
More dimensions - 101:05:19
More dimensions - 201:07:18
Second order rescaling - 101:07:27
Rescaling weights01:09:04
Second order rescaling - 201:09:18
Practical issues01:09:20
Standard solutions01:10:01
Attention01:12:07
Simple things we can do01:13:22
Training multilayer networks Initialization01:15:21
Random weight initialization01:15:38
The simplest two-layer net - 101:16:19
The simplest two-layer net - 201:16:57
The simplest two-layer net - 301:17:42
Initialization01:19:13