Knowingo
The cover of Perceptrons

Perceptrons

Marvin Minsky, Seymour A. Papert

A rigorous guide to what single layer learning machines can and cannot represent, showing why architecture, prior structure, and scaling costs matter alongside convergence.

Get the app Read on the web

6 parts, about 18 minutes to read.

At a glance

Author
Marvin Minsky, Seymour A. Papert
Parts
6 parts
Reading time
about 18 minutes

What the summary covers

  1. What a perceptron can learn2 min
  2. A weighted decision from fixed features2 min
  3. Order, symmetry, and difficult combinations3 min
  4. Geometry exposes the limits and the exceptions5 min
  5. Representation can be costly even when possible2 min
  6. Beyond one layer and toward scale4 min

More books like this