# perceptron

> algorithm for supervised learning of binary classifiers

**Wikidata**: [Q690207](https://www.wikidata.org/wiki/Q690207)  
**Wikipedia**: [English](https://en.wikipedia.org/wiki/Perceptron)  
**Source**: https://4ort.xyz/entity/perceptron

## Summary
The perceptron is an algorithm designed for the supervised learning of binary classifiers. Invented in 1957 by Frank Rosenblatt, it functions as a type of feedforward neural network in which connections between nodes do not form a cycle. It serves as a fundamental unit in artificial intelligence, notably highlighted in the publication "The perceptron: a probabilistic model for information storage and organization in the brain."

## Key Facts
- **Inception**: The perceptron was invented in 1957.
- **Inventor**: It was created by Frank Rosenblatt, an American psychologist notable in the field of artificial intelligence.
- **Classification**: It is an instance of an algorithm and a subclass of the feedforward neural network.
- **Function**: The algorithm is used for supervised learning of binary classifiers.
- **Structural Definition**: As a feedforward network, its connections between nodes do not form a cycle.
- **Key Publication**: It is described by the source "The perceptron: a probabilistic model for information storage and organization in the brain."
- **Related Entity**: The multilayer perceptron is a related class described as a feedforward neural network with multiple fully connected layers.

## FAQs
### Q: Who invented the perceptron?
A: The perceptron was invented by Frank Rosenblatt, an American psychologist recognized for his work in artificial intelligence. He introduced the concept in 1957.

### Q: What type of algorithm is a perceptron?
A: It is an algorithm for supervised learning specifically designed for binary classifiers. Structurally, it is classified as a feedforward neural network where connections between nodes do not form cycles.

### Q: How does the perceptron relate to a multilayer perceptron?
A: The multilayer perceptron is a related class of technology characterized as a feedforward neural network with multiple fully connected layers.

## Why It Matters
The perceptron represents a critical historical and structural milestone in the development of artificial intelligence and machine learning. Invented in 1957 by Frank Rosenblatt, it bridged the gap between psychology and computer science, offering a "probabilistic model for information storage and organization in the brain." As the foundational subclass of feedforward neural networks, it established the groundwork for more complex architectures, such as the multilayer perceptron and kernel perceptron. Its existence as a non-cyclic network architecture laid the rules for how nodes connect in many modern AI systems, making it a primary reference point in the study of neural networks.

## Notable For
- Being the first feedforward neural network architecture, invented in 1957.
- Bridging the fields of psychology and computer science through its inventor, Frank Rosenblatt.
- Serving as the basis for the "probabilistic model for information storage and organization in the brain."
- Being a subclass of acyclic networks (feedforward) which distinguishes it from recurrent neural networks.
- Inspiring related architectures like the kernel perceptron and multilayer perceptron.

## Body
### Definition and Classification
The perceptron is an algorithm utilized for the supervised learning of binary classifiers. It is formally classified as a subclass of the **feedforward neural network**, a category of artificial neural networks defined by the property that connections between nodes do not form a cycle. It is recognized as a distinct instance of an algorithm within computer science taxonomies.

### Origins and Inventor
The concept of the perceptron was realized in **1957**. Its discoverer and inventor was **Frank Rosenblatt**, an American psychologist (born July 11, 1928) who was also noted as a computer scientist and neuroscientist. Rosenblatt's work established the perceptron as a model for how information might be stored and organized biologically, though it operates as a computational algorithm.

### Related Architectures
The perceptron serves as the parent or predecessor to several other classes and architectures:
- **Multilayer Perceptron**: A class of feedforward neural network distinguished by having multiple fully connected layers.
- **Kernel Perceptron**: A class of algorithm related to the original perceptron.

### Descriptions and Identifiers
The algorithm is widely documented across various academic and encyclopedic sources. It is described by the specific work *"The perceptron: a probabilistic model for information storage and organization in the brain"* as well as general encyclopedias such as the Armenian Soviet Encyclopedia and *Perceptrons* (book). It holds the GND ID 4173941-3 and is categorized under the Commons category "Perceptrons."

## References

1. Integrated Authority File
2. Freebase Data Dumps. 2013
3. Quora
4. National Library of Israel
5. [OpenAlex](https://docs.openalex.org/download-snapshot/snapshot-data-format)