Skip to content

Latest commit

 

History

History
50 lines (31 loc) · 1.8 KB

README.md

File metadata and controls

50 lines (31 loc) · 1.8 KB

Artificial Neural Network Learning Rules

Classification is made in 2-dimensional space with artificial neural networks learning rules.

Learning Factors

  • Initial Weights
  • Steepness of the Activation Function
  • Learning Constant
  • Momentum Method
  • Necessary Number of Hidden Neurons
  • Network Architectures Versus Data Representation
NOTE : Additionally, GUI indicators such as loss charts and period counter negatively affect runtime efficiency.

- SingleCategory SingleLayer Neural Network

- discrete (Perceptron Learning Rule)
- continuos (Delta Learning Rule)
DISCRETE CONTINOUS
SinglePerceptron SingleDelta

- MultiCategory SingleLayer Neural Network

- Discrete (Perceptron Learning Rule) 
- Continous (Delta Learning Rule)
DISCRETE CONTINOUS
MultiCategoryPerceptron MultiCategoryDelta

MultiLayer Neural Network

- Error Back Propagation
MULTILAYER
MultiLayer

License: MIT