Basics of Deep Learning

BY
HOOR FATIMA
 Course Contents
Module 1:-Basics of Deep learning

Deep learning architectures, Convolutional Neural Networks, Neurons in Human Vision-

The Shortcomings of Feature Selection-Vanilla Deep Neural Networks, Don’t Scale-Filters
and Feature Maps-Full Description of the Convolutional Layer-Max Pooling-Full
Architectural Description of Convolution Networks-Closing the Loop on MNIST with
Convolutional Networks-Image Preprocessing Pipelines Enable More Robust Models-
Accelerating Training with Batch Normalization-Building a Convolutional Network for
CIFAR-10-Visualizing Learning in Convolutional Networks.

Module 2 Memory Augmented Neural Networks

Neural Turing Machines-Attention-Based Memory Access-NTM Memory Addressing

Mechanisms-Differentiable Neural Computers-Interference-Free Writing in DNCs-DNC
Memory Reuse-Temporal Linking of DNC.
Reinforcement Learning: Deep Reinforcement Learning Masters Atari Games, What Is
Reinforcement Learning?-Markov Decision Processes (MDP)-Explore Versus Exploit-
Policy versus Value Learning-Pole-Cart with Policy Gradients-Q-Learning and Deep Q-
Networks
 Course Contents
Module 3
Web Analytics – Basics – Traditional Ways – Expectations – Data Collection – Clickstream
Data – Weblogs – Beacons – JavaScript Tags – Packet Sniffing – Outcomes data–Competitive
data–Search Engine Data.
Qualitative Analysis – Customer Centricity – Site Visits – Surveys – Questionnaires –
Website Surveys – Post visits – Creating and Running- Benefits of surveys – Critical
components of successful strategy.

Module 4

Web Analytic concepts – URLS – Cookies – Time on site – Page views – Understand
standard reports – Website content quality – Navigation reports (top pages, top destinations,
site overlay). – Search Analytics – Internal search, SEO and PPC – Measuring Email and
Multichannel Marketing - Competitive intelligence and Web 2.0 Analytics–Segmentation–
Connectable reports.
 Course Contents
Module 5

Qualitative Analysis: Customer Centricity – Site Visits – Surveys – Questionnaires –

Website Surveys – Post visits – Creating and Running- Benefits of surveys – Critical
components of successful strategy. Web Analytic concepts – URLS – Cookies – Time on
site – Page views – Understand standard reports – Website content quality – Navigation
reports (top pages, top destinations, site overlay). – Search Analytics – Internal search,
SEO and PPC.

Google Analytics: Analytics - Cookies - Accounts vs Property - Tracking Code Tracking

Unique Visitors - Demographics - Page Views & Bounce Rate Acquisitions Custom
Reporting.
Suggested texts and reference materials

1. Phil Kim, “MATLAB Deep Learning: With Machine Learning, Neural Networks and
Artificial Intelligence”, First Edition , Apress, 2017.
2. Nikhil Buduma, Nicholas Locascio, “Fundamentals of Deep Learning: Designing Next-
GenerationMachineIntelligenceAlgorithms”,O'ReillyMedia,2017.
Suggested texts and reference materials
3. Avinash Kaushik, “Web Analytics 2.0: The Art of Online Accountability and
ScienceOfCustomerCentricity“,1stedition,Sybex,2009.
4. Michael Beasley, “Practical Web Analytics for User Experience: How Analytics
canhelpyouUnderstandyourUsers”,MorganKaufmann,2013.
5. Bing Liu, “Web Data Mining: Exploring Hyperlinks, Content, and Usage Data”, 2nd
Edition,Springer,2011.
JustinCutroni,“GoogleAnalytics”,O’Reilly,2010. 6. Eric Fettman, Shiraz Asif, Feras
Alhlou , “Google Analytics Breakthrough”, John Wiley&sons,2016.
 AGENDA FOR TODAYS LECTURE

 What is Intelligence?

 What is Artificial Intelligence?

 What is Machine Learning?

 Limitations of Machine Learning

 What is a need to Deep Learning?

 What is Deep Learning?

 Deep Learning applications

What is “Intelligence”
 It can be described as the ability to gather information, and
to retain it as a knowledge to be applied in some situation.
 The idea come from the Latin word: intelligo which means
“to choose between two objects”.

What is “Intelligent behavior”

 Intelligent behavior is generally decided by how a person
performs in response to a problem the answer to which is
not known
What do we need to be an “intelligent”

 The capacity for logic

 Understanding
 Self awareness
 Learning
 Emotional knowledge
 Reasoning
 Planning
 Creativity
 Problem solving
 What is Artificial Intelligence
It is the capability of a machine to mimic intelligent human behavior.

AI is accomplished by Outcomes of this study is

studying human brain thinks used as a basis of
and how human brain learn, developing intelligent
decide and work while trying software and systems
to solve a problem
Applications of artificial intelligence
 Finance : Banks use artificial intelligence systems to organize operations, invest in
stocks, and manage properties. In August 2001, robots beat humans in a simulated
financial trading competition.
 Medicine: A medical clinic can use AI systems to organize bed schedules, make a staff
rotation, and provide medical information. AI has also application in fields of
cardiology (CRG), neurology (MRI), embryology (sonography),complex operations of
internal organs etc.
 Social Sites: Facebook wants to analyze the way people communicate with one another
so that it can add new features to its services or even automatically remove offensive
posts that might occur when a high-profile celebrity like Angelina Joli posts a selfie.
 Robotics: Robots are physical agents that perform tasks by manipulating the physical
world . Robots are manufactured as hardware. The control of robot is AI (in the form of
software agent) that reads data from the sensors decides what to do next and then
directs the effectors to act in the physical world.
Applications of artificial intelligence

 Heavy Industries : Huge machines involve risk in their manual maintenance and
working. So in becomes necessary part to have an automated AI agent in their
operation.
 Education: AI researchers have created many tools to solve the most difficult problems
in computer science
What does machine need to be “intelligent”

 Perception- understanding images, audio etc.

 Reasoning-answering questions from data

 Planning-inferring the required steps to reach a goal

 Natural language processing- understanding human language

 Ability to learn from past experiences

Machine learning (ML)

 It is a technique which is used to achieve Artificial Intelligence. It uses

machines the ability to “learn from data” without being explicitly
programmed.

Drawback of Machine learning

The main drawback of machine learning is the lack of self awareness.
 Deep Learning
 It is a technique for implementing Machine Learning.

 It uses Artificial Neural Networks

 Artificial Neural Network are inspired by our biological

neurons i,.e, interconnections between the neurons

Machine Learning
Input Feature Classification Output
extraction

Deep Learning
Input Classification Output
Why do we need deep learning
 Huge amount of data

 Complex problems

 Feature extraction

What is deep learning

 Handle huge amount of structured and unstructured data

 Complex operations problems solved
 It is a learning technique that learn features and tasks directly from
data, whereas data maybe images, test or sound
Where deep learning algorithms can be applied

 Medical field

 Robotics

 Self driving cars

 Translation