Neural network
- 1. V.SAKTHI PRIYA INFO TECH NADAR SARASWATHI COLLEGE OF ARTS ANDSCIENCE
- 2. Neural networks are simplified models of the biological neuron system. Neural network is a massively parallel distributed processing system highly interconnected neural computing elements. Neural network refer to learning process in trainning and to solve a problem using on inference.
- 3. Artificial neural system(ANS) Artificial neural network(ANN) Simply Neural networks(NN)
- 4. Neuron is a highly complex structure viewed as a highly interconnected network of simple processing elements. Biological neuron is termed as artificial neuron . In this model basis of artificial neural network.
- 5. X1 weights Thresholding output x2 w3 x3 Summation of weighted inputs THRESHOLDING UNIT xn w1 w2 . . wn
- 6. x1,x2,x3 ……..xn are the n inputs. w1,w2,…..,wn are the weights to be input links. The total input I received by the artificial neuron. I=w1x1+w2x2+………..+wnxn. n wi xi i=1
- 7. Artficial neural network is a define as a dataprocessing system consisting a large number of simple highly interconnected processing elements in artificial neuron. It can be represented using a directed graph. A graph G is on order two tuples(V,E). V represent set of vertices, E represent set of edges.
- 8. The graph is directed call is a directed graph or digraph. v5 v3 v4v2 v1 e1 e2 e3 e4 e5
- 10. Single layer feedforward network have two layers. SLFFN INPUT LAYER OUTPUT LAYER
- 11. The input layer neurons receive the input signal. The output layer neurons receive the output signal. The input layer transmits the signals to the output layer. It is called as single layer feedforward network.
- 12. w11 xi: Inputneuron w12 yj: Outputneuron w1n wij: weights w21 . w22 . . wn1 . . Wn2 wnm wnm Input layer output layer x1 y1 x2 y2 xn ym
- 13. Multi layer feedforward network has including multiple layers. MLFFN INPUT LAYER OUTPUT LAYER HIDDEN LAYER
- 14. Input layer and output layer have one or more intermediary layers. It is called ‘hidden layer’. The computational units of the hidden layer are know as the hidden neurons or hidden units. Input hidden layer weights: The input layer neurons are linked to the hidden layer neurons and the weights are link on these are called ‘input hidden layer weights’.
- 15. Input hidden layer weights: The input layer neurons are linked to the hidden layer neurons and the weights are link on these are called “input hidden layer weights”. Hidden output layer weights: Hidden layer neurons are linked to the output layer neurons and weights are reffered to the “hidden output layer weight”. L-M1-M2-N.
- 16. x2 x1 y1 xn ym z1 z2 z3 zn w11 w12 w13 w1n v11 v1m v21 v2m vl1 vlm Xi:input neurons Yi:hidden neurons Zk:output neurons Vij:input hidden Layer weight Wjk:output hidden Layer weights
- 17. Recurrent network is differ from feedforward network architecture. There is atleast one feedback loop. There could also be neurons with self- feedback links. Example: The output of a neuron is fed back into itself as input.
- 19. The NNS exhibit mapping capabilities,that is, they can map input patterns to their associated output patterns. The NNS posses the capability to generalize.thus they can predict new outcomes from past trends. The NNS are robust system and are fault tolerant.they can,therefore recall full patterns from incomplete,partial or noisy patterns. The NNS can process information in parallel at high speed and in a distributed manner.
- 20. NEURAL NETWORK LEARNING SUPERVISED LEARNING(ERROR BASED) ERROR CORRECTION GRATIENT DESENT LEAST MEAN SQUARE BACKPROP AGATION STOCHASTIC UNSUPERVI SED LEARNING HEBBIN COMPETITIVE REINFORCE D LEARNING( OUTPUT BASED)
- 21. Learning methods in neural network can be broadly classified into three basic types. supervised Un supervised Reinforced