Artificial Neural Network

1. Artificial Neural Network
Ali Raza
University of Gujrat
Pakistan

2. Defination
• ANNs is biological inspired model
• Simulate the way in which human brain is
working
• ANN gather the knowledge by detecting
patterns and relationship in data
• It learnt from experience not from
programming

3. How it is formed
• ANN is formed from hundreds of Single Units,
Artificial Neurons or Processing elements,
connected with coefficients
• Coefficient: Which constitute the neural
structure and are organized in layers
• Power of Neural Computations comes from
connecting Neurons.

4. ANN
• Artificial Neurons or Processing Elements
Consist of
i. Weighted Input
ii. Transfer Function
iii. One Output

5. Cont…
• Behavior of Neural Network is Determined by
transfer function of its Neurons, by
architecture itself, by learning rules
• Weights are the adjustable parameters, in that
sense, Artificial Neural Network is the
parameterized system

6. Cont…
• Weighted sum of inputs constitutes the
activation of neurons
• The input signal pass through the transfer
function and produce output of neuron
• Transfer function produce Non linearity of the
function

7. Cont..
• The inter-unit connections are optimized until
the error in predictions is minimized and the
network reaches the specified level of
accuracy.
• Once the network is trained and tested it can
be given new input information to predict the
output

8. Cont…
• Many types of neural networks have been
designed already and new ones are invented
every week but all can be described by the
transfer functions of their neurons, by the
learning rule, and by the connection formula
• ANN represents a promising modeling
technique, especially for data sets having non-
linear relationships which are frequently
encountered in pharmaceutical processes

9. Cont…
• In term of model specification ANN not
Required the information about the data
source but the weights must be estimated,
they require large number of training set
• ANN can combine and incorporated both
literature based and experimental based data
to solve problem

10. Application
• Pattern Recognition
• Prediction and modeling
• Face Recognition
• Iris Recognition
• Optical Path Finding
etc