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

---

- MultiCategory SingleLayer Neural Network

- Discrete (Perceptron Learning Rule) 
- Continous (Delta Learning Rule)

DISCRETE	CONTINOUS

---

MultiLayer Neural Network

- Error Back Propagation

MULTILAYER