Autoplay
Autocomplete
Previous Lesson
Complete and Continue
311. Neural Network Visualization
Introduction
Welcome
Make a sketch (2:43)
Add dimensions (5:07)
Lay out the visualization
Code up the constants
Build a parameter dictionary I: Figure (3:14)
Build parameter dictionary II: Input (1:26)
Build parameter dictionary III: Layers (2:52)
Build parameter dictionary IV: Error image (1:21)
Create the Figure (3:19)
Set Figure colors (5:16)
Set Figure resolution (2:37)
Build the visualization
Calculate image size based on height (3:14)
Calculate image size based on width (4:34)
Find the node image size (2:48)
Find the gap between layers (1:17)
Find the gap between nodes (2:06)
Find the error image position (1:56)
Add the input image
Add a random input image (2:15)
Remove ticks and labels (3:52)
Outline the image (0:53)
Refactor add_image_axes() (1:03)
Refactor add_filler_image() (2:01)
Sidebar: The challenge of writing a good plotting language
Add the rest of the images
Add the first node in the first layer (1:13)
Add the whole first layer (3:17)
Add the rest of the layers (1:52)
Add the output image (2:32)
Add the error image (1:05)
Use a real filler image (0:38)
Add connections
Build a list of layers of Axes (6:49)
Connect the input image and the first node in the first layer (4:07)
Connect the input image and the whole first layer (4:58)
Connect each layer with the next (1:02)
Spread out connection points (2:16)
Make connections curvy (3:01)
Vary connection width (3:23)
Differentiate positive connections from negative (1:56)
Wrap up
Next steps (1:19)
Resources
Add a random input image
Lesson content locked
If you're already enrolled,
you'll need to login
.
Enroll in Course to Unlock