ARTIFICIAL NEURAL networks.

1
ARTIFICIAL NEURAL networks.

• Presented by
• Mudit Misra
• Deveshri Srivastava
• Richa Sharma
• Neerja Gupta

2
WHAT IS A NEURAL NETWORK ?

• An Artificial Neural Network (ANN) is an
  information processing paradigm that is inspired
  by the way biological nervous systems, such as
  brain, process information.
• It comprises of a large number of highly
  interconnected processing elements (neurons)
  working in unison to solve specific problems.

3
WHY USE NEURAL NETWORKS ?

• Adaptive Learning
• Self Organization
• Real Time Operation
• Fault Tolerance via Redundant Information Coding

4
NEURAL NETWORKS VERSUSCONVENTIONAL COMPUTERS

• Neural Networks do not need a pre-defined
  algorithm to execute an instruction.
• The data used by Neural Networks need not be very
  precise.
• The information in a Neural Network is processed
  by constantly changing patterns of activity.

5
The biological inspiration

• Neuron consists of a branching input structure-
  dendrites, a cell body, and a branching output
  structure-axon.
• Each neuron can propagate electrochemical
  signals.

Dendrites
Soma (cell body)
Axon
6
HOW THE HUMAN BRAIN LEARNS
dendrites
axon
synapses

• The information transmission takes place at the
  synapses.
• Neuron , when activated fires an electrochemical
  signal along the axon, which travels through
  synapses to the other neuron.
• The strength of the synaptic connections is
  responsible for the learning process.

7
from human neurons to artificial neurons

• Artificial Neurons receive and provide
  information in the form of spikes.
• The below model is known as the McCullough-Pitts
  model.

x1 x2 x3 xn-1 xn
Output
Inputs
8
THE BASIC ARTIFICAL MODEL

• The artificial neuron receives one or more input
  signals, sums these, and produces an output .
• Activation of neuron weighted sum of inputs
  threshold.
• The output is produced after passing the sum
  through a non-linear function known as an
  activation or transfer function.

9
HOW SHOULD NEURONS BE CONNECTED TOGETHER ?

• If a network is of any use, there must be an
  input and an output.
• There are hidden neurons also that play an
  internal role in the network.
• The input , hidden and output neurons need to be
  connected together.

10
TYPES OF NEURAL NETWORK

• Feed forward Network
• .
• Recurrent Network.

11
APPLYING A NEURAL NETWORK TO SOLVE A PROBLEM

• A solution to the problem is defined by the way
  the network works and the way they are trained.
• They are used to infer some unknown information
  from known information.
• The condition to get a solution is that there
  must be a relationship between the proposed known
  input and unknown output.

12
TYPES OF TRAINING

• Supervised learning
• Unsupervised learning

13
IMPLEMENTATION AND FUTURE TECHNOLOGY

• Neural networks mimics the brain, they have shown
  much promise in so-called sensory processing
  task.
• Neural networks can perform as well as humans.
• Neural network have the ability to learn from a
  set of examples and generalize this knowledge to
  new situation, they are excellent for work
  requiring adaptive control system.
• They can sustain damage and still function
  properly.
• Neural networks are currently a hot research area
  in medicine.