Lecture No.

01 AI-7 , SEN-5
th th

Course: Introduction to Data Science

Instructor: Dr. Maryum Nisar

12/15/2024 1
 Course Objective
Objective 1:
 To understand complete process of data science.
Objective 2:
 To get deep understanding of classification algorithms and regression methods,
feature extraction, and segmentation techniques.
Objective 3:
 To understand data analysis methods, various evaluation measures, and ensemble-
based methods.

https://covid19.uthm.edu.my/wp-content/uploads/
2020/04/Data-Science-from-Scratch-First-Principles-with- 2
Python-by-Joel-Grus-z-lib.org_.epub_.pdf
 Course Outline
Data Science Complete Process
 Data Collection
 Data Preprocessing Process
 Data Preprocessing Techniques
Regression
 Types of Regression
 Regression Techniques
 Regression Analysis
Feature Engineering
 Feature Extraction Techniques
 Feature Selection Techniques
https://covid19.uthm.edu.my/wp-content/uploads/
2020/04/Data-Science-from-Scratch-First-Principles-with- 3
Python-by-Joel-Grus-z-lib.org_.epub_.pdf
 Course Outline
Learning Paradigms
 Supervised Learning
 Unsupervised Learning
 Semi-Supervised Learning
Classification Algorithms
 Common Classification Algorithms
 Ensemble-Based Algorithms
Evaluation Measures
 Metrics to Evaluate Model Performance

https://covid19.uthm.edu.my/wp-content/uploads/
2020/04/Data-Science-from-Scratch-First-Principles-with- 4
Python-by-Joel-Grus-z-lib.org_.epub_.pdf
 Course Outline
Clustering Techniques
 Segmentation Techniques
Dimensionality Reduction Techniques
Data Analysis Techniques
Reinforcement Learning
Natural Language Processing
Data Imbalance
Techniques to Handle Imbalanced Data
Deep Learning
https://covid19.uthm.edu.my/wp-content/uploads/
2020/04/Data-Science-from-Scratch-First-Principles-with- 5
Python-by-Joel-Grus-z-lib.org_.epub_.pdf
Course Weight Breakdown
Assessment Instruments with Weights
 Homework, Quizzes, Midterms, Final, Assignments etc.)
 Theory
o Quizzes 10%
o Assignment 10%
o Midterm 30%
o Presentation/Other Activities 10%
o Final 40%

12/15/2024 6
 Course book
Text & Reference Books
 Focus will be on topics rather
Chapters of any text
 Many Textbooks are available in
market
o Data Science from Scratch by Joel Grus
Copyright © 2019 Joel Grus. All rights
reserved. Printed in the United States of
America.

https://covid19.uthm.edu.my/wp-content/uploads/
2020/04/Data-Science-from-Scratch-First-Principles-with- 7
Python-by-Joel-Grus-z-lib.org_.epub_.pdf
 Semester Project
All students in pairs will choose a real time dataset based on the
data you will be finding the insights by applying AI(ML, DL and etc.)
Datasets can be extract from genuine data centers such as kaggal,
IBM data center, and google data center and many more
First assignment will be the description of your data, its parameters
why you have considered the chosen one what technique you are
going to apply

12/15/2024 8
 Introduction to Data Science

12/15/2024 9
 Lecture Contents
Course introduction
 Introduction to Data Science
 List of best jobs
 Data Science and Artificial Intelligence
 Data science is not machine learning
 Data science is not statistics
 Data science is not big data
 Data Science as a UNIFIER
 Applications of data science

12/15/2024 10
 “Data! Data! Data!” he cried impatiently. “I can’t make bricks without
clay.” —Arthur Conan Doyle

12/15/2024 11
Introduction to Data Science
Data Science
 Data science is the application of computational and statistical techniques to
address or gain insight into some problem in the real world
 Data science = statistics +
data processing +
machine learning +
scientific inquiry +
visualization +
business analytics +
big data + …

12

12/15/2024 12
 Data Scientists

13
https://www.ranker.com/crowdranked-list/best-jobs-in-the-world
https://www.panelplace.com/blogs/top-8-coolest-jobs-world 13
 Data Scientists

 Data Scientist
 The hottest Job of the 21st Century
 They find stories, extract knowledge
 They are not reporters

12/15/2024 14
 Lists of Best Jobs

15
https://www.ranker.com/crowdranked-list/best-jobs-in-the-world
https://www.panelplace.com/blogs/top-8-coolest-jobs-world 15
 Lists of Best Jobs

https://www.glassdoor.com/List/Best-Jobs-in-America-LST_KQ0,20.htm
16
https://money.usnews.com/careers/best-jobs/rankings/the-100-best-jobs
 Data science and AI

“Data science produces insights. Machine

learning produces predictions”

12/15/2024 17
 Data science is not machine learning
Machine learning involves computation and statistics, but has not
(traditionally) been very concerned about answering scientific
questions
Machine learning has a heavy focus on fancy algorithms

12/15/2024 18
 Data science is not statistics
“Analyzing data computationally,
to understand some phenomenon
in the real world, you say? …
that sounds an awful lot like statistics”
Not many statistics courses have a lecture
on e.g. web scraping, or a lot of data
processing more generally
Plus, statisticians use R, while data scientists
use Python ... clearly these are completely different fields
https://www.linkedin.com/pulse/data-science-statistics-paulo-
19
rios-jr-
 Data science is not big data
Sometimes, in order to truly understand and
answer your question, you need massive
amounts of data
But sometimes you don’t
Don’t create more work for yourself than
you need to

Gendered language in professor reviews. Image

20
source: http://benschmidt.org/profGender
 DATA SCIENCE AS A UNIFIER

12/15/2024 21
 DATA SCIENCE APPLICATION
 Fraud detection
 Investigate fraud patterns in past data
 Early detection is important
o Before damage propagates
o Harder than late detection
 Precision is important
o False positive and false negative are both bad
 Real-time analytics

12/15/2024 22
 Cont.
 Recommender systems
 The ability to offer unique personalized service
 Increase sales, click-through rates, conversions, …
 Netflix recommender system valued at $1B per year
 Amazon recommender system
drives a 20-35% lift in sales annually
 Collaborative filtering at scale

12/15/2024 23
 Cont.
Predicting why patients are being readmitted
 Reduce costs
 Improve population health
 Find the “why” behind specific populations
being readmitted
 Data lakes of multiple data sources
 Investigate ties between readmission and
socioeconomic data points, patient history, genetics

12/15/2024 24
 Cont.
Moneyball
 How to build a baseball team on a very
low budget by relying on data
 Sabermetrics: the statistical analysis of baseball
data to objectively evaluate performance
 2002 record of 103-59 was joint best in MLB
o Team salary budget: $40 million
 Other team: Yankees
o Team salary budget: $120 million

12/15/2024 25
 References
https://dsg.uwaterloo.ca/CDSW/slides/Data%20Science%20Present
ation.pdf
https://github.com/MicrosoftLearning/Data-Science-Essentials/
blob/master/Slides/Module%201%20-%20Intro%20to%20Data
%20Science.pptx
Data Science, Computer Science and Mathematical Sciences College
of Engineering Tennessee State University

12/15/2024 26
Upcoming lecture
IDE for data science
Exploration of data
Data and types

12/15/2024 27
 Thank You !
12/15/2024 28