Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 9

A PRESENTATION ON

“Deep Neural Network”


By
Nikhil Sunil Patil
Class: T. Y. B. Tech –B
PRN: 201101131

Subject Name: Machine Learning (ML)


TA 1- Presentations

Department of Computer Engineering


R. C. Patel Institute of Technology, Shirpur
An Autonomous Institute
Content :

 What is Deep Neural Network?

 How Does a Deep Neural Network Work?

 Advantages of Deep Neural Networks

 Applications of Deep Neural Networks

 Limitations of Deep Neural Networks

 Conclusion

1
What is Deep Neural Network?

 Deep Neural Networks


(DNNs) are a type of
artificial intelligence (AI)
that uses a large number of
layers of artificial neurons to
process data. This type of AI
is used to solve complex
problems that require
advanced analysis and
understanding.

Figure : Deep Neural Network


 DNNs are used in a variety
of fields, from computer
vision and natural language
processing to robotics and
autonomous vehicles.

2
How Does a Deep Neural
Network Work?

 Deep Neural Networks are


composed of layers of
artificial neurons, which are
connected together to form a
network. The neurons are
organized into input, hidden,
and output layers. Data is fed
into the input layer, where it is
processed and passed through
the hidden layers. Finally, the
output layer produces a result.

 Neurons are grouped into three Figure : Working of DNN’s


different types of layers:
a) Input layer
b) Hidden layer
c) Output layer
3
Advantages of Deep Neural Networks :

• DNNs can learn complex patterns and relationships in data.

• DNNs can achieve high accuracy in various applications.

• DNNs can adapt to new data without significant changes to the model.

• DNNs can automatically extract relevant features from raw data.

• DNNs can handle large datasets and complex architectures.

• DNNs can be applied to a wide range of applications.

4
Applications of Deep Neural Networks :

Figure : Applications of DNN’s

5
Limitations of Deep Neural Networks :

 Overfitting: DNNs can easily overfit on the training data, leading to poor generalization to
new, unseen data. Regularization techniques such as dropout and weight decay can help
address this issue.

 Computational complexity: DNNs can be computationally expensive to train, especially


when dealing with large datasets or complex architectures. This can limit their scalability in
certain applications.

 Data requirements: DNNs require large amounts of labeled data to be effective. In situations
where data is scarce or expensive to obtain, this can be a limiting factor.

 Vulnerability to adversarial attacks: DNNs can be vulnerable to adversarial attacks, where


small perturbations to the input can cause the model to make incorrect predictions.

 Limited applicability: While DNNs have shown impressive results in various applications,
they may not be the best choice for all problems. In some cases, simpler models or alternative
approaches may be more effective or efficient.
4
Conclusion

 Deep Neural Networks are a powerful tool for solving complex problems. They are able
to learn complex patterns from data and make accurate predictions. However, they are
computationally expensive and can be difficult to interpret.

 Despite these limitations, DNNs have a wide range of applications and are becoming
increasingly popular in many fields.

6
References:
[1] Han Kamber,“Data Mining Concepts and Techniques”, Morgann Kaufmann
Publishers. (Reference Book)

[2] Peter Harrington, “Machine Learning In Action”, DreamTech Press ( Text Book)

You might also like