ANN

ANN

1. Benefits of ANN

   • Useful for pattern recognition
   • Analyse large quantities of data
   • Can Work with incomplete data
   • Adaptable to New environments

2. Machine Learning

   • Biological neural networks
     • Composed of cells neurons
     • Network has several 1000 neurons
     • Brain has a collection of neural networks
     • Partitioned into groups called networks
   • ANN used to automate complex decision making
   • learn from past experience
     • First Gen PDAs hand writing recognition
   • learn from past experience
   • mimics the processing capabilities of the human brain
   • massive parallel processing”
   • ANN identifies patterns

3. ANN Processing of info

   • Summation
   • Transformation

4. Processing information in ANN

   • Neuron (Processing Element PE)
     • Inputs
       • Input correspondent to a single attribute loan age, income etc
     • Weights
       • By repeated adjustments of weights networks learn
     • Outputs
       • The output can be final or fed into other neurons
       • Solution to the problem ie YES or NO

5. ANN Applications

   • Speech recognition
   • Character Recognition
   • Data Transmission
   • Financial predictions