Question

The advantages and disadvantages of Design Hamming, Hopfield and Perceptron neural network

Answer 1

Hamming Code :

1. The Hamming code is the set of collecting the error-correction code and it can be used to detect and also find the error in the code

2. It can be processed and can be done when the data is moved or stored from the sender part to the receiver part.

3. This is developed by the R.W hamming for his name only the process is arrived as the Hamming code and it is used for the error correction.

Advantages of Hamming Code :

1.it Provides the protection of the data in bits whenever it is with the smaller redundancy and that is reduced.

2. And the other important aspect is having the bigger hamming code is the size of the coder and decoder.

3. It is extremely effective on the different types of networks whenever the data streams can be given in a single-bit error.

Disadvantages of Hamming Code :

1. It only can solve the single bit and the correction code.

2. It can solve up to the single bits only.

Hopfield :

1. It is a type of single layer and it contains the single or one / more connected neurons.

2. This Hopfiles network is commonly used for the auto association and performing the optimization tasks.

3. This Hopfield model can have the one inverting and the one non-inverting outputs.

4. Here the output of each neuron that should be the input of the other neurons but it is not the input of its self.

5. This Hopfield Network is up of two types they are

1. Discrete Hope Network and

2. Continuous Hopfield Network.

Advantages of Hopfield Network :

1.it is good for the content type of the address memory and for solving some kind of optimisation problems.

2. It does not have any type of network learning algorithms.

3.here the patterns are simply stored by using and settings their weights to the lower the network energy.

The disadvantage of Hopfield Network :

1. This Hopfield network is the type and it needs a lot of data in order to check the process.

2. Here the Training time should be very long it takes longer than usual.

Perceptron Network :

1. The perceptron is a concept that is introduced in early 1950 by the minsky and the pepert.

2. It explains about the classify of the linearly separable input of the sets.

3. In Perceptron concepts, we are having the perceptron convergence theorem that is introduced by Rosenblatt in the year 1962.

4. This perceptron can mainly deal with the AND, OR, NOT, XOR

5.The Multi-Layer perceptron deals with the backpropagation networks.

Advantages of Perceptron Network :

1.the Multilayer perceptron network that can be trained by using the backpropagation algorithm

2.this is helpful for performing any kind of mapping between the Input and the Output.

Disadvantages of Perceptron Network:

1. It is requiring the labelled data of the training.

2. Here the learning time does not scale well in the Perceptron Neural Network.

3. It can be struck by the poor local optimum.