1. Modelleerimine ja Juhtimine Tehisnärvivõrgudega
Identification and Control with artificial neural networks
Eduard Petlenkov,
Eduard Petlenkov, 2013

2. Artificial neuron Input vector: Vector of weight coeffitients:
Weighted sum
Nonlinear element
Input vector:
Vector of weight coeffitients:
Weighted sum:
Eduard Petlenkov, 2013

3. Biological neuron and biological neural networks
Eduard Petlenkov, 2013

4. Sigmoid functions are having an "S" shape (sigmoid curve)
Activation functions (1)
OUT=f(NET)
Sigmoid functions are having an "S" shape (sigmoid curve)
1 Logistic function
2 Hyperbolic tangent
Eduard Petlenkov, 2013

5. Types of artificial neural networks
Feedforward
Feedback
internal feedback
external feedback
Structure:
inputs
etalon
Supervised learning
Selflearning (unsupervised)
Training/ learning:
inputs
Training criteria
Eduard Petlenkov, 2013

6. Interpritation of results
How to use artificial neural networks
Network
Data preprocessing
Interpritation of results
Inputs
Outputs
Eduard Petlenkov, 2013

7. Feedforward neural networks and multilayer perceptron
Feedforward networks are networks in which an output of a neuron can be connected only with an input of a next layer neuron.
- Input layer
- Output layer
- Hidden layer
- Weighting coefficients, where
is the number of the neuron’s input
“from each to each”
- neuron’s number in the layer
- number of the layer
Eduard Petlenkov, 2013

8. Mathematical function of a two-layer perceptron
- Activation function of the hiddent layer neurons;
- Activation function of the output layer neurons.
Eduard Petlenkov, 2013