perceptron

Binary classification unit

A perceptron is the simplest type of artificial neural network—a single node that takes multiple inputs, applies weights to them, and produces a binary output.

Developed in the 1950s, it was an early attempt to model how neurons fire in the brain.

Though limited on its own, perceptrons became building blocks for more complex neural networks: The course started with perceptrons before moving to deep learning architectures.

architecture

neuron

Impulse-conducting cell

weights · bias · input · output · layer · node · unit · linear · classifier · step

learning

training

Developing abilities through practice

supervised · delta · update · rule · epochs · error · correction · iteration · linearly · separable

algorithm

gradient

Rate of change across space

function · backpropagation · convergence · iteration · update · margin · decision · boundary · XOR · problem

history

Mark

Common masculine name

  • IFirst-person subject pronoun
  • CornellIvy League institution
  • 1958Mid-century calendar year

AI · winter · neural · networks · cybernetics · cognitive · science · pioneers · revival

Opposites
network depth/complexity

perceptrontransformer

Voltage-altering electrical device

GPT

learning paradigm

perceptronexpert system

Software emulating specialist reasoning

decision tree · knowledge graph · human brain

historical era in AI

perceptronlarge language model

AI trained on vast text

ChatGPT · GPT-4

Family

deception, perception, perceive, perceptive, misconception, prospective, perceptible, introspection, perceptions

Familiarity
Top 1,000,000
Profile
Used worldwide · Timeless · Abstract · Rated G