9/14/2019

Data Science
Dr. Muzammil Khan

Assistant Professor
Department of Computer & Software Technology

Office # 4

Evaluation
 Evaluation Criteria
 Total Marks 100% (100)
 Final Term Exam 50%
 Mid Term Exam 30%
 Assignments + Presentations + Quizzes 10%
 Term Paper (Major Assignment) 10%

 Recommended Readings
 Data Science, Theories, Models, Algorithms and Analytics
 By Sanjiv Ranjan Das
 Data Science from Scratch (First Principal with Python)
 By Joel Grus
 Internet as a best source
 Research Articles
Digital Image Processing
2

1
 9/14/2019

Evaluation (Cont...)
 Policies
 Late assignments may be accepted with marks reduction.
There will be a 10% reduction for assignments submitted up
to 24 hours late.
 Students who have copied assignments or whose
assignments have been copied will both be given a zero.
 Plagiarism is not acceptable. Anyone found to be guilty of
plagiarism, the assignment will be marked as zero in that
assignment.
 Quizzes may be unannounced or announced (depends on
your response).

 Term paper is compulsory as semester project

Digital Image Processing
3

Data Science (DS) Course Outline

 Introduction to Data Science
 Statistical Inference
 Data Extraction, Wrangling (preparing) and Exploration
 Introduction to
 Machine Learning
 Data Mining
 Classification Techniques
 Unsupervised Learning / Clustering Techniques
 Recommender Systems
 Text / Web Mining
 Natural Language Processing
 Deep Learning
Data Science
4

2
 9/14/2019

Chapter 1

Introduction to
Data Science

Data Science
5

In this Chapter
 Data
 Big Data
 Big Data Challenges
 Introduction to DS
 Its Applications
 DS Core Components
 Use Cases Examples
 Data Scientists
 Introduction to Hadoop & R
 R & Hadoop Integration
 Machine Learning with Hadoop
 Some important terms
Data Science
6

3
 9/14/2019

Data & Its Sources

 A lots of Data Sources exist
 Lots of data is being collected and warehoused
 Even, streaming continuously

 For Example
 Web data,
 E-commerce,
 Financial transactions, bank/credit transactions,
 Online trading and purchasing,
 Social Network,
 Etc…

Data Science
7

How Big it is ?
 Still growing…

Data Science
8

4
 9/14/2019

Huge Data Centers (Millions of Servers)

Data Science
9

Big Data
 Big Data is any size data that is
 Expensive to manage &
 Hard to extract knowledge from

 Focus of Big Data !!!

 on 3 V’s

Data Science
10

5
 9/14/2019

3 V’s of Big Data (another perception)

Data Science
11

5 V’s of Big Data

Data Science
12

6
 9/14/2019

Big Data Analytics

 Why Analytics ?

Data Science
13

Big Data Challenges

 The main problem is;

Data Science
14

7
 9/14/2019

Big Data Challenges (Cont…)

Data Science
15

Data Science
 Data Science is
 “An area that manages, manipulates, extracts, and interprets
knowledge from tremendous amount of data”

 A multidisciplinary field of study with goal to address the

challenges in big data

 So,
 Data Science is the science which uses
 Computer science, statistics and machine learning,
visualization and human-computer interactions
 To collect, clean, integrate, analyze, visualize, interact with
data to create data products.

Data Science
16

8
 9/14/2019

Data Science (Cont…)

 Turning Data into Data Product
 A data product is a deliverable from
 Data Discovery
 Data Prediction
 Data Service
 Data Recommendation
 The ultimate data products are
 Knowledge
 Intelligence
 Wisdom
 Decision

 Data science principles apply to all data – Big and Small

Data Science
17

Data Science is Multidisciplinary

Data Science
18

9
 9/14/2019

Data Science “A Bigger Picture”

Data Science
19

Hype Cycle (Gartner’s 2014)

Data Science
20

10
 9/14/2019

Data Science “Applications”

Data Science
21

Data Science “Applications” (Cont…)

 Transaction Databases  Recommender systems (NetFlix),
Fraud Detection (Security and Privacy)

 Wireless Sensor Data  Smart Home, Real-time Monitoring,

Internet of Things

 Text Data, Social Media Data  Product Review and

Consumer Satisfaction (Facebook, Twitter, LinkedIn), E-
discovery

 Software Log Data  Automatic Trouble Shooting (Splunk)

 Genotype and Phenotype Data  Epic, 23andme, Patient-

Centered Care, Personalized Medicine

Data Science
22

11