You are on page 1of 7

ARTIFICIAL INTELLIGENCE

EL40102
CONTENTS

֍ Biological Neuron
֍ Artificial Neuron
֍ Activation Function
֍ Single Layer Perceptron
֍ Multi Layer Perceptron
֍ Back Propagation Algorithm
֍ Recurrent Neural Network
֍ Adaptive Resonance Theory

Asutosh Patnaik, Asst. Prof., Dept. of EEE, CGU,


10-08-2022 2
Odisha
Unit – 1 Lecture 5
Objective of Today’s Lecture
Understand the concept of Backpropagation Algorithm

Asutosh Patnaik, Asst. Prof., Dept. of EEE, CGU,


10-08-2022 3
Odisha
Backpropagation Algorithm
Backpropagation (backward propagation of errors)

Backpropagation is an algorithm used in AI to fine-tune the weights based on the error obtained in the
previous iterations and improve the accuracy of an ANN’s outputs.

Asutosh Patnaik, Asst. Prof., Dept. of EEE, CGU,


10-08-2022 4
Odisha
Backpropagation Algorithm

Forward pass:
Start by propagating the data inputs to the input
layer, go through the hidden layer(s).
Measure the network’s predictions from the output
layer, and
Finally calculate the network error based on the
predictions the network made.

Asutosh Patnaik, Asst. Prof., Dept. of EEE, CGU,


10-08-2022 5
Odisha
Backpropagation Algorithm
Example:
if the correct output is 4 and the network’s prediction
is 1.3, then the absolute error of the network is 4-
1.3=2.7.
Note that the process of propagating the inputs from
the input layer to the output layer is called forward
propagation.
Once the network error is calculated, then the
forward propagation phase has ended, and
backward pass starts.

Asutosh Patnaik, Asst. Prof., Dept. of EEE, CGU,


10-08-2022 6
Odisha
Backpropagation Algorithm
 Inputs X, arrive through the preconnected
path
 Input is modeled using real weights W.
The weights are usually randomly
selected.
 Calculate the output for every neuron from
the input layer, to the hidden layers, to the
output layer.
 Calculate the error in the outputs
Error = Actual Output – Desired Output

 Travel back from the output layer to the


hidden layer to adjust the weights such
that the error is decreased.
Keep repeating the process until the desired
output is achieved

Asutosh Patnaik, Asst. Prof., Dept. of EEE, CGU,


10-08-2022 7
Odisha

You might also like