Plate modeling

As explained in Causal AI

Plate modeling is a probabilistic-model visualization technique that groups repeated variables into nested graphical substructures.

Plate modeling visualizes repeated variables in a probabilistic machine-learning or causal DAG while preserving their lower-level dimensional detail. A rectangle, or plate, groups repeated variables into a subgraph; a number records how many entities repeat, and nested plates represent repetition within repetition. section 5.1.2

Image pixels and training data

An image represented by 784 pixels need not appear as 784 separate DAG nodes. A plate can group pixels Xj, with j indexing each pixel; another plate can represent the repeated images in a training set. section 5.1.2

Figure 5.7 A plate model representation of the causal DAG. Plates represent repeating variables, in this case 28 × 28 = 784 pixels. X_j is the j^th pixel.

Use the plate-model diagram to inspect how 28 × 28 pixels become indexed repeated variables within a causal DAG. section 5.1.2

How Causal AI covers it

How deep it goes

Explains plates, repetition counts, indexing, and nesting, but stops short of estimation or implementation details.

Examples

Uses 28 × 28 image pixels, handwritten and typed digits, and repeated training images.

What you can do after

Evaluate/understand compact graphical representations of repeated image variables and nested datasets.

How it connects

Connects causal DAGs, probabilistic machine learning, image pixels, MNIST, and TMNIST.

liveBookmanning.com

Read Causal AI on liveBook Get the book at manning.com

Related concepts

• plate notationuses · Plate notation is a graphical notation that represents repeated variables or subgraphs with enclosing rectangles and repetition indices.

This explainer was generated by Manning's knowledge graph from the text of Causal AI. The section links point to the parts of the book it draws on.