Page 166 - Ai Book - 10

through proper testing so that one can calculate the efficiency and performance of the model. Hence, the model
is tested with the help of Testing Data and the efficiency of the model is estimated on the basis of the parameters
such as accuracy, recall and precision,or the F1 score.

NEURAL NETWORK
A neural network is essentially a way of organizing machine learning algorithms to perform a certain task. The
most impressive aspect of a neural network is that once trained, it can learn on its own without any intervention.

To put it simply, an Artificial Neural Network can be defined as a computing system made up of simple, highly
interconnected processing elements which process information by their dynamic state response to external
inputs.

Working of Neural Network

A neural network has great potential to solve problems for which the dataset is very large such as images,
patterns etc. In simple words, the larger neural networks are able to produce better results with larger amounts
of input data as compared to traditional machine learning algorithms.
A neural network is the complex structure in AI systems which is made up of artificial neurons . This network is
able to take multiple inputs to produce a single and meaningful output. In general, a neural network consists of
an input and output layer with one or multiple hidden layers within. The basic architecture of a neural network
is shown in the figure given below:
Input layer Hidden layer Output layer

Input # 1

Input # 2

Output
Input # 3

Input # 4

Here, input layer nodes receive only one value or input at one time and send it to all the nodes available in the
hidden layer. The hidden layer nodes perform specific functions on the incoming data and pass the processed
information to the output layer node. You should remember that no processing would take place at the input
layer due to its passive nature. Each node available in the hidden layer has its own machine learning algorithm.
The execution of the algorithm starts when it receives data from the input layer. After that, processed information
is then passed to the subsequent hidden layer if available. It is important to note that a neural network may
contain one or more than one hidden layer. The number of hidden layers depends upon the complexity of a
function for which it was configured.
In the last stage, the last hidden layer passes the processed information to the node of the output layer. As
similar to input layer, output layer cannot make any modification in the acquired information.
The concept of neural network have been widely used in pattern recognition and classification tasks due to the
following features:

u ANNs are capable of learning complex nonlinear input-output relationships.

40
40