In this tutorial, you will learn:

Contents

## Back Propagation Neural Network

Before we learn Back Propagation Neural Network (BPNN), let’s understand:

## What is Artificial Neural Networks?

A neural network is a group of connected I/O units where each connection has a weight related with its computer programs. It assists you to build predictive models from large databases. This model expands upon the nervous system. It assists you with directing picture understanding, human learning, computer speech, and so on

## What is a neural network?

In information technology (IT), an artificial neural network (ANN) is a system of hardware and/or software patterned after the operation of neurons in the human brain. ANNs – likewise called, just, neural networks- are an assortment of deep learning technology, which additionally falls under the umbrella of artificial intelligence, or AI.

## What is Backpropagation?

Backpropagation is the essence of neural network training. It is the strategy for adjusting the weights of a neural network dependent on the error rate acquired in the previous epoch(i.e., iteration). Appropriate tuning of the weights allows you to diminish error rates and make the model dependable by expanding its generalization.

Backpropagation in neural network is a short form for “in backward propagation of errors.” It is a standard technique for training artificial neural networks. This strategy helps calculate the gradient of a loss function with respect to all the weights in the network.

## How Backpropagation Algorithm Works

The Back propagation algorithm in neural network computes the gradient of the loss work for a single weight by the chain rule. It effectively computes one layer at a time, in contrast to a native direct computation. It computes the gradient, however it doesn’t define how the gradient is used. It generalizes the computation in the delta rule.

- Inputs X, show up through the preconnected path
- Input is modeled using genuine weights W. The weights are generally 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

ErrorB= Actual Output – Desired Output

- Travel back from the output layer to the hidden layer to change the weights such that the error is diminished.

Continue to repeat the process until the desired output is accomplished

## Why We Need Backpropagation?

Most prominent benefits of Backpropagation are:

• Backpropagation is fast, simple and easy to program

• It has no parameters to tune apart from the numbers of input

• It is a flexible technique as it doesn’t need earlier information about the network

• It is a standard technique that by and large functions well

• It does not need any special mention of the features of the function to be learned.

## What is a Feed Forward Network?

A feedforward neural network is an artificial neural network where the nodes never form a cycle. This sort of neural network has an input layer, hidden layers, and an output layer. It is the first and easiest sort of artificial neural network.

## Types of Backpropagation Networks

Two Types of Backpropagation Networks are:

• Static Back-propagation

• Recurrent Backpropagation

### Static back-propagation:

It is one sort of backpropagation network which delivers a mapping of a static input for static output. It is useful to solve static classification issues like optical character recognition.

### Recurrent Backpropagation:

Recurrent Back propagation in data mining is fed forward until a proper value is accomplished. From that point forward, the error is computed and propagated backward.

The main difference between both of these strategies is: that the mapping is fast in static back-propagation while it is nonstatic in recurrent backpropagation.

## History of Backpropagation

• In 1961, the basics concept of ceaseless backpropagation were derived in the context of control theory by J. Kelly, Henry Arthur, and E. Bryson.

• In 1969, Bryson and Ho gave a multi-stage dynamic system optimization technique.

• In 1974, Werbos sated the possibility of applying this principle in an artificial neural network.

• In 1982, Hopfield brought his idea of a neural network.

• In 1986, by the effort of David E. Rumelhart, Geoffrey E. Hinton, Ronald J. Williams, backpropagation acquired recognition.

• In 1993, Wan was the first individual to win an international pattern recognition contest with the assistance of the backpropagation strategy.

## Backpropagation Key Points

• Simplifies the network structure by elements weighted links that have minimal impact on the prepared network

• You need to study a group of input and activation values to develop the relationship between the input and hidden unit layers.

• It assists with evaluating the effect that a given input variable has on a network output. The information acquired from this analysis ought to be addressed in rules.

• Backpropagation is particularly useful for deep neural networks working on error-prone projects, like picture or speech recognition.

• Backpropagation takes advantage of the chain and power rules allows backpropagation to function with any number of outputs.

## Best practice Backpropagation

Backpropagation in neural network can be clarified with the assistance of “Shoe Lace” analogy

**Too little tension =**

• Not enough constraining and very lose

**Too much tension =**

• Too much constraint (overtraining)

• Taking an excessive amount of time (relatively slow process)

• Higher likelihood of breaking

**Pulling one lace more than other =**

• Discomfort (bias)

## Disadvantages of using Backpropagation

• The actual performance of backpropagation on a specific issue is reliant upon the input data.

• Back propagation algorithm in data mining can be quiet sensitive to noisy data

• You need to use the matrix-based methodology for backpropagation rather than mini-batch.

## Summary

• A neural network is a group of associated it I/O units where each association has a weight associated with its computer programs.

• Backpropagation is a short form for “backward propagation of errors.” It is a standard technique for preparing artificial neural networks

• Back propagation algorithm in machine learning is fast, simple and easy to program

• A feedforward BPN network is an artificial neural network.

• Two Types of Backpropagation Networks are 1)Static Back-propagation 2) Recurrent Backpropagation

• In 1961, the essentials idea of continuous backpropagation were derived in the context of control theory by J. Kelly, Henry Arthur, and E. Bryson.

• Back propagation in data mining simplifies the network structure by eliminating weighted links that minimally effect on the prepared network.

• It is particularly useful for deep neural networks working on error-prone projects, such as picture or speech recognition.

• The biggest drawback of the Backpropagation is that it can be sensitive for noisy data.

Thanks for reading! We hope you found this tutorial helpful and we would love to hear your feedback in the Comments section below. And show us what you’ve learned by sharing your projects with us.

## Frequently Asked Questions

## Back Propagation Neural Network

A neural network is a group of connected I/O units where each connection has a weight related with its computer programs.

## What is a neural network?

In information technology (IT), an artificial neural network (ANN) is a system of hardware and/or software patterned after the operation of neurons in the human brain.

## What is Backpropagation?

Backpropagation is the essence of neural network training. It is the strategy for adjusting the weights of a neural network dependent on the error rate acquired in the previous epoch(i.e., iteration). Appropriate tuning of the weights allows you to diminish error rates and make the model dependable by expanding its generalization.

## How Backpropagation Algorithm Works

The Back propagation algorithm in neural network computes the gradient of the loss work for a single weight by the chain rule. It effectively computes one layer at a time, in contrast to a native direct computation. It computes the gradient, however it doesn’t define how the gradient is used.

## Why We Need Backpropagation?

• Backpropagation is fast, simple and easy to program

• It has no parameters to tune apart from the numbers of input

• It is a flexible technique as it doesn’t need earlier information about the network

• It is a standard technique that by and large functions well

## What is a Feed Forward Network?

A feedforward neural network is an artificial neural network where the nodes never form a cycle. This sort of neural network has an input layer, hidden layers, and an output layer.

## Types of Backpropagation Networks

**Static back-propagation:**

It is one sort of backpropagation network which delivers a mapping of a static input for static output.

**Recurrent Backpropagation:**

Recurrent Back propagation in data mining is fed forward until a proper value is accomplished.

## History of Backpropagation

• In 1961, the basics concept of ceaseless backpropagation were derived in the context of control theory by J. Kelly, Henry Arthur, and E. Bryson.

• In 1969, Bryson and Ho gave a multi-stage dynamic system optimization technique.

• In 1974, Werbos sated the possibility of applying this principle in an artificial neural network.

• In 1982, Hopfield brought his idea of a neural network.

## Backpropagation Key Points

• Simplifies the network structure by elements weighted links that have minimal impact on the prepared network

• You need to study a group of input and activation values to develop the relationship between the input and hidden unit layers.

## Best practice Backpropagation

Backpropagation in neural network can be clarified with the assistance of “Shoe Lace” analogy

**Too little tension =**

• Not enough constraining and very lose

**Too much tension =**

• Too much constraint (overtraining)

## Disadvantages of using Backpropagation

• The actual performance of backpropagation on a specific issue is reliant upon the input data.

• Back propagation algorithm in data mining can be quiet sensitive to noisy data