Scribd
Upload
5 views

Deep Learning Unit1

Deep learning is a machine learning technique that builds hierarchical models called deep neural networks to improve predictive accuracy across various applications like computer vision and speech recognition. Prerequisites for learning deep learning include programming (preferably in Python or R), statistics, calculus, linear algebra, probability, and data science. The future of deep learning holds promise across industries but faces challenges related to transparency and bias.

Uploaded by

Rahul Yadav
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
5 views

Deep Learning Unit1

Deep learning is a machine learning technique that builds hierarchical models called deep neural networks to improve predictive accuracy across various applications like computer vision and speech recognition. Prerequisites for learning deep learning include programming (preferably in Python or R), statistics, calculus, linear algebra, probability, and data science. The future of deep learning holds promise across industries but faces challenges related to transparency and bias.

Uploaded by

Rahul Yadav
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 126

Deep Learning

ELECTIVE-VIII (BTCOE705 (B))


What is Deep Learning
• Deep learning is a machine learning technique used to obtain a more accurate
predictive model of your data, which can then be used to predict with higher
accuracy how users will behave.

• It works by building a hierarchical model called a deep neural network.

• It consists of multiple processing layers, each layer consists of multiple


neurons that interact with each other.

• It is used in a broad range of applications, including computer vision (self-


driving cars), speech recognition (virtual assistants), and in recommendation
systems.
Text Books
• Ian Goodfellow and Yoshua Bengio and Aaron Courville, "Deep
Learning", 1st Edition, MIT Press

• Reference Books:
1. Raúl Rojas, Neural Networks: A Systematic Introduction, 1996.
2. Christopher Bishop, Pattern Recognition and Machine Learning, 2007
Course Outcomes
1. Describe the basics and architecture of deep neural networks
Level: Recall

2. Analyze the various components of Deep Neural networks


Level: Analyze

3. Elaborate different approaches of regularization for error reduction in


deep learning models
Level: Understand

• 4. Describe different types of convolutional neural networks


• Level: Understand
What are the prerequisites to learn deep learning?
• You need to have knowledge of how machine learning works.

• To have a basic understanding of computer programming.

• You don't need to be an expert in programming (there are already many


languages specialized in machine learning), but you will need to know the
basics of how a computer works and how it uses data to make decisions.

• Knowledge of some basic math will be very useful.

• Since machine learning is based on statistics and probability, learning some


statistics and probability will help you understand machine learning better.
Deep Learning Prerequisites

• 1. Programming

• Programming is the fundamental requirement of deep learning.

• You can’t perform deep learning without using a programming language.

• Deep learning professionals use Python or R as their programming


language because of their functionalities and effectiveness.
Python Programming
• Defining and calling functions, using positional and keyword
parameters
• Dictionaries, lists, sets (creating, accessing, and iterating)

• Conditional blocks and conditional expressions


• String formatting
• Variables, assignment, basic data types
2. Statistics

• Statistics refer to the study of using data and its visualization.

• It helps In gaining information from the raw data you have.

• In statistics, you plot charts, create graphs, and understand relations


between different data points.

• It also helps you gain insights from samples of data and classifying the
available information in different segments according to your
requirements.
Statistics
• mean, median, outliers, and standard deviation
• ability to read a histogram
3. Calculus

• Calculus forms the basis for many machine learning algorithms.

• In deep learning, you’ll be building models according to the features


present in your data.

• Calculus will help you in using those features and


making the model accordingly.
Calculus
• Concept of a derivative (you won't have to actually calculate
derivatives)
• Gradient or slope
• Partial derivatives (which are closely related to gradients)
4. Linear Algebra

• Probably one of the most important deep learning prerequisites is linear


algebra.

• Linear algebra deals with matrices, vectors, and linear


equations.
• It focuses on the representation of linear equations in vector spaces.

• The linear algebra will help you in building models of various sorts
(classification, regression, etc.), and it is another building block for
numerous concepts of deep learning.
Algebra
• variables, coefficients, and functions
• linear equations such as
• logarithms, and logarithmic equations such as
• sigmoid function

• Linear algebra
• matrix multiplication
• tensor and tensor rank (multi-dimensional arrays with a uniform
type)

• Trigonometry
• tanh (discussed as an activation function; no prior knowledge
needed)
5. Probability
• Probability is a branch of mathematics that focuses on describing how
likely an event can occur or how possible it is valid through numbers.

• The probability of any event ranges from 0 to 1, where 0 indicates


impossibility, and 1 represents absolute certainty.

• In ML and deep learning, you have to build models for predictive analysis.
• You have to train them to predict specific outcomes.

• That’s why probability is an essential subject to study for a deep learning


student.
6. Data Science

• Data science is the field of analyzing and using data to gain valuable
insights.

• Data science helps companies in making predictions about customer behavior,


sales, and market trends.

• As a deep learning professional, you must be familiar with various concepts of


data science as you’d have to build models that handle data.

• Knowing deep learning will help you in using data to get the desired results.
What is the future of deep learning?
• Deep learning is used across industries ranging from medical to e-commerce.
• In the medical industry, deep learning is used to identify cancerous growths
in MRIs.
• In e-commerce, deep learning is used to determine which advertisements
and products to display to customers.
• The two major challenges facing deep learning technology today are
transparency and bias.
• Transparency is the ability for a human to understand the reasoning behind
a machine-made decision.
• Bias is when a machine is consistently favoring a certain outcome.
• Because of these challenges, the future of deep learning technology is
uncertain.
• 1950: Alan Turing published “Computer Machinery and Intelligence” which
proposed a test of machine intelligence called The Imitation Game.
McCulloch Pitts Neuron
Artificial Neuron
Artificial Neuron
Biological vs. Artificial Neuron
Artificial Neuron

You might also like