Artificial Neural Networks

 

Artificial Neural Network is one of the domain of Deep Learning which can simply be 

understood as replica of functioning of human brain Artificially.

Now lets understand What is Artificial Neural Network ?

Artificial Neural Network is a Technology which is based on the functioning

of human brain. It is composed of large number of highly connected  processing 

elements to solve a  particular  problem. 

How does Artificial Neural Network actually works ?

ANN actually imitate the working of biological neurons . ANN is made up of 

several artificial neurons which  recieve inputs in the from of  an internal

weighting  system  and generates outputs according to the information 

presented to it.It uses a learning rule which is known as back propagation

of error to generate exact and accurate output.

Types of Artificial Neural Networks:

Well  basically there are two types of ANN:

1. FeedForward Artificial Neural Network:

    In this type when one neuron sends an information  to  another  neuron 

    there is no feedback sytem meaning it has fixed input and  output.

2. FeedBack Artificial Neural Network: 

    In this type of ANN  there is feedback loop between message transmission

    between one neuron to other neuron 

  Usage of Machine Learning in Artificial Neural Network:

  

      1. Supervised Learning:

          The learning  can  be thought  of  as  teacher  supervising learning  process .

          It uses an  algorithm which  knows  the output according   to  the  input

          data.The algorithm makes prediction  the  training  data  andis  corrected

          by  the teacher.The learning stops when  the algorithm  reaches  it 

          appropriate output or result.  

  2.Unsupervised  Learning:

       In this type of learning  there  are  no  correct  answer is not given there is 

       no teacher. Algorithms are left to  their  own  to  their  own to discover and

       present  their  output or result

 

3. Reinforcement Learning:

     In  this learning  agents  have to  face a very  complex situation  to  achieve

     a goal. The computer employs trial  and  error   to  come  up  with a 

     solution  to  a problem. 

  Conclusion:

  One of the key  features of Artificial  Neural  Networks is that it gives output

  according  to  the  data  feeded  ie  it does not  have  capability  to take decisions on

  its own however researchers  are continuosly making key developments in ANN

  to come up with bigger innovations and development  in  the field  of ANN  .