Applied Areas
Popular Course in this category
Health Care Retail E-commerce Online recommendations
Finance Health Care Retailing Machine learning
Tracking price changes Better customer service and delivery systems
Examples
Siri, Google Maps and Google Search, etc.
Artificial Intelligence
Stock Exchange Prediction
Image Recognition, Image Compression, and Search engines etc.
Conclusion Machine Learning vs. ANN Machine Learning falls under the same field of Artificial Intelligence, wherein Neural Network is a subfield of Machine Learning.
Machine learning serves mostly from what it has learned, wherein neural networks are deep learning that powers the most human-like intelligence artificially.
Neural networks or deep learnings are the next evolution of machine learning. It explains how a machine can make their own decision accurately without any need for the programmer telling them so.
Q1. What is a neural network in the context of computing? Ans. Artificial neural network (ANN) refers to a collection of connected
computational units or nodes (called neurons) along with biologically inspired
computer programs designed to stimulate the way in which the natural neural
networks of human brain process information.
Q2. How does a neural network work?
Ans. In a neural network, information transfers from input layer nodes to
connected layer(s), which after required processing and decision-making
reach to the final layer as output. Neural networks are initial trained with
some input data and then they keep learning with every new input and
feedback.
Q3. How do ‘neurons’ talk to one another in a human neural network
Ans. The communication between two neurons happens at the synapse, the
small gap between two neurons or between a neuron and a target cell, like a
muscle or a gland. The information signals are passed via electro-chemical
processes and can be thought of as a simple electrical impulse that travel from
axon to dendrite.
Q4. What is the role of an activation function in a neural network?
Ans. An activation function in a neural network defines how the weighted
sum of the input is transformed into an output from a node or nodes in a layer
of the network.5
Q5. How do neural networks improvise and learn?
Ans. Like human brain, when we provide an input or feedback about the
expected output, the brain registers it and modifies its way to reach at that
conclusion. Based on the received feedback about the difference from thecorrect output, the neurons in ANNs make changes in the weights or bias so
that they can reach to the correct input.
Q6. List the features of neural network.
Ans. We can list the features of neural networks as follows:
Neural networks have been developed to mimic the structure and working of
human brain.
Neural networks evolve and automatically learn with each input and each new
attempt.
Neural networks can work with big data sets.
The neural networks employ machine learning techniques to function and
evolve.
Let Us Revise
Artificial neural network (ANN) refers a collection of connected
computational units or nodes (called neurons) along with biologically inspired
computer programs designed to simulate the way in which the natural neural
networks of human brain process information.
The message transfer from a neuron is called action potential.
In an ANN, a node (or neuron) is the artificial equivalent to neuron. It consists
of a set of weighted inputs (dendrites), an activation function (soma) and one
output (axon). Information all through multiple layers of connecting neurons
before it is transformed in the form on an output.
A Neural Network is divided into multiple layers (hidden layers in between
output and input layers).
Each layer of ANN consists of several artificial neurons called nodes.
Each neuron node has to perform a specific took and pass the information
to the next layer.
An activation function in a neural network defines how the weighted sum
of the input is transformed into an output from a node or nodes in a layer
of the network.
Providing feedback about the difference from the correct output is known as
back propagation.
After initial training, Neural networks keep updating their internal
mechanism of reaching to an output with every new feedback received.
Neural networks can be applied to all popular types of Al models such
as Regression model (a rule based model), clustering model (A machine
learning approach model), Classification model (a rule bused model).
In supervised learning, ANNs learn through labelled data of tasks with
known outcomes.
Unsupervised learning, ANNs learn through unlabelled data of tasks
with unknown outcomes
Some popular applications of neural networks are Face recognition,
Speech recognition, Self-driven vehicles, medical diagnosis and many
others.
Deep learning is the most advanced form of machine learning where
there are many hidden computational layers in a neural network (deep
neural network) enabling an algorithm to powerfully analyse the input
data.1. In an artificial neural network, the equivalent of neuron in human
neural network is called:
(a) Node
(b) Neuron
(c) Neurode
(d) All of these
2. In an artificial neural network, the equivalent of dendrites in human
neural network is called:
(a) Dendrites
(b) Weighted inputs
(c) Artendrites
(d) All of these
3. In an artificial nueral network, the output is made available through:
(a) Neuron
(b) Synapse
(c) Axon
(d) Activation function
4. In an artificial neural network, the equivalent of soma in human
neural network is called:
(a) Neuron
(b) Synapse
(c) Axon(d) Activation function
5. Providing feedback about the difference from the correct output is known:
(a) Reverse input (b) Update (c) Data input (d) Back propagation
6. Most advanced form of artificial intelligence is:
(a) Machine learning (b) Neural network (c) Deep learning (d) All of these
7. Which of the following is incorrect: (a)
In rule based approach, the relationship or patterns in data are defined by the developer
(b) Decision tree looks like an upside-down tree (c) Pixel it activity is an example of how computers see images, process them and classify them
(d) In learning based approach, the relationship or patterns in data are defined by the developer.
No comments:
Post a Comment