The Modern Mathematics of Deep Learning

Circular window operation

Penguin learning project

The project uses 342 complete Palmer penguin records: 240 training-pool and 102 held-out rows. The selected training rows determine feature scaling and model parameters. Source CSV

This project trains weights with softmax, unlike the homepage lab, which uses nearest species averages. Same penguin measurements; different classification methods. Predictions can differ.

Frames and sparse signals · Encode or reconstruct signals with frames and dictionaries.

The project scales bill and flipper measurements using training-row means and scales; its feature vector has two entries.

Consider this concept alongside the project; its source assumptions may differ. These observations do not prove a theorem.

Input: 12 × 2; Held-out evaluation: 102 · #58 · 40.6 mm, 193 mm. Known answer: Adélie; Prediction: Adélie · [≈0.7383, ≈0.2444, ≈0.01737].

Used in these source sections: 6.3 Taking apart and rebuilding

A circular window takes consecutive values with wraparound and forms their dot product with v.

The source uses this identity to relate window matrices and convolutions.

Source formula example

\langle e^{(i)},\mathbb H_n(x)v\rangle=\sum_{j\in H}x_{i+j}v_j

exact · The result is exact for these finite toy inputs, apart from ordinary computer rounding.

Symbols

$G$: Cyclic positions in the input.
$H$: Window positions.
$v$: Window weights.

Source formula example

This source example uses its own inputs and assumptions. Editing it does not change the project.

Edit a value or the code, then run the example. This is a finite demonstration, not a proof of the paper’s theorem.

Zero-based position

This example’s data flow

You choose

input.position = 3

Initial values in code

x = [1, 2, 3, 4]
v = [1, 2]

Calculated here

Result expression: { window: v.map((_, j) => x[(i + j) % x.length]), value: v.reduce((sum, weight, j) => sum + x[(i + j) % x.length] * weight, 0), }

i = input.position Uses: input.position

Returns

Example result

window: array (2 items): [4, 1]
value: 6

JavaScript

Read numbers from input and return the result. JavaScript basics.

Result

Enable JavaScript to run this example.

Technical statement

The finite example uses zero-based indices with wraparound. It evaluates the middle sum in the source identity.

Source: Circular window identity