CSCI E-96
Data Mining for Business
Spring Term 2024
Course Information
CRN: 26599
Section Number: 1
Format: Flexible Attendance Web Conference
Credit Status: Graduate
Credit Hours: 4
Course Description: This course introduces non-mathematical business professionals
to data science principles widely used in today's corporations. Quantitative methods
affect many of today's interactions for business leaders, students, and consumers.
Emphasis is placed on practical uses and case studies utilizing data to inform business
decisions rather than theoretical or complex mathematics. Case study topics include
understanding customer demand, marketing, new market forecasting, revenue
projections, and data mining to improve decisions. Learning goals include quantitative
business application, basic programming, algorithm development, and process workflow.
The course highlights methods that business leaders and data scientists have found to
be the most useful. It introduces the basic concepts of R for data mining. This course is
for students who want an introduction to how data science improves business outcomes.
Prerequisites: Students need to bring laptops to class to use for exercises.
Instructor Information & Office Hours
Edward Kwartler
Email: [email protected]
Kalle Georgiev
Preferred Pronouns: she/her
Email: [email protected]
Section Meetings
Additional lab sessions to support the case study assignments will be hosted by the TAs.
The dates and times are announced within the class discussion forum "Ed Discussion"
and are recorded.
Course Goals / Learning Outcomes
If you stay engaged in the course and complete the suggested code challenges and
assignments:
You will be able to think systematically about how data is used to make business
decisions. This objective will be accomplished through the use of ideas from statistics,
economics, and computer technology and using business-related case studies.
Students will learn how to implement a variety of popular data mining algorithms in R (a
free and open-source software) to tackle business problems and identify opportunities.
This course will help introduce the basics of R in data mining.
As a business leader, you will acquire the skill of applying data science concepts within
business domains to improve decisions and learn how data scientists approach projects.
As a data scientist, you will acquire practical applications of data mining methods that are
used in many of today’s most successful organizations as well as understand what
business stakeholders expect of data scientists.
Mode of Attendance & Participation Policy
This is a flexible attendance course, which means you can choose to: (1) attend class
live over Zoom (synchronous option); or (2) watch the class recording afterward
(asynchronous option). You do not need to commit to the same mode of attendance for
the whole semester.
If you are attending live over Zoom:
Class meetings take place over Zoom. Because they involve active participation,
discussion, and dialogue, you are expected to attend all class meetings. Please arrive on
time. You should attend Zoom meetings with a functional web-camera and microphone,
prepared with materials needed, to engage thoughtfully, and with your camera on. You
may turn off your camera for occasional interruptions or momentarily for privacy.
You will also need the most up-to-date Zoom client installed on your computer to join
class. Please participate from a safe and appropriate environment with appropriate
clothing for class. Participating while traveling or in a car is not permitted. In addition,
please do not join class via mobile phone or web browser.
If you are participating asynchronously:
You are expected to watch the class recording and complete any assignments before the
next live class meets.
Please be sure to review important information on Student Policies and Conduct.
Grading & Grade Definitions
Grading
A course grade will be assigned based on student performance on case studies and
applicable homework assignments (undergraduate).
Assignments are accepted up to 12 hours late. Any work submitted after the deadline but
before 12 additional hours will be penalized 10% of the total weight of the assignment.
After 12 hours no late submissions will be accepted under ANY circumstances. Pupils
are expected to manage their own time and submit their work accordingly. Failure to
submit submissions through the University approved portal by the assignment deadline
will be considered late and not accepted. Submissions to any other location will not
be accepted.
Graduate Level Assignments
Skills Assessment: 0%: Complete the provided but unfinished R script.
Case I 25%: Retail Store Transaction EDA
Case II 25%: Banking Case
Case III 40%: Hotel/Lodging Booking Case
Assessment 10%: Time Series Forecast TSA Passengers
Extra Credit: Homework II Visualization in R 1% of total grade
Undergraduate Level Assignments
Skills Assessment: 0%: Complete the provided but unfinished R script.
Case I 25%: Retail Store Transaction EDA
Case Case II 25%: Banking Case
Homework I: 10% Intro to R script - listed in class repository
Homework II: 20% Visualization in R script - listed in class repository
Homework III: 20% Obtain 2 AI/Data Ethics related articles (https://incidentdatabase.ai is
a good resource). Perform the following tasks for each article.
Use ChatGPT to summarize the article
Critique the summarization as appropriate, inappropriate, missing relevant facts,
creating or citing information from outside the article etc. in a single paragraph.
Cite the gaps in understanding from the technology you note to provide a more
nuanced understanding
Write 1-3 paragraph(s) WITHOUT GPT with your personal reflection on the use or
misuse of the technology cited in the article. In the paragraph suggest ways to
mitigate or monitor to protect against the issue within the article.
Case Work Product & Expectations
Each case will have a description and specific instructions provided through the course
github repository.
Each student will work on case studies individually. Each case will have the following
work artifacts:
Maximum ~10min recorded slide presentation uploaded to youtube (provide the
link in your Canvas submission), embedded as a voiceover in the slides or shared
in a similarly appropriate manner to Canvas. The presentation will outline the
business problem, the insights identified, describe the data and the
outcomes/recommendations satisfying the case
Slide presentation uploaded to Canvas (pptx file for example), R Script(s)
supporting the creation of any visuals, models or recommendations made during
the presentation.
At least one case requires predictions be made on new data. For example,
"identify the top 100 prospective customers". If mentioned in the specific case
instructions, please score, identify the appropriate observations, save results in a
CSV and submit that file as well.
A Written supplemental: Submit something to represent the entirety of your
presentation including the data, process, findings, and implications. Thus it’s a
professional report, anything less than a professionally written and organized report
will be considered sub-optimal. Amazon for example doesn’t use PowerPoint and
instead uses “6 pagers” to make business recommendations, as such some
organizations prefer written information over presentations. The use of external and
verifiable sources is expected to add context and support any component of the
paper. The minimum is 2 pages maximum is 5. Double spaced and 12 point
font.
Essentially all supporting material to satisfy the case including scripts, visuals,
presentation slides, data tables and/or written document will need to be turned in for
review.
Case Evaluation
Organization – Was the presentation well organized?
Delivery – Was the content delivered clearly and persuasively with the audience in
mind?
Code Documentation – Was the data mined to support the conclusion?
Written Supplemental – Is the information clear and supported in narration and code? Did
the information satisfy the case problem?
Data Mining & Modeling Process – Overall, as a complete portfolio of work, is the topic
interesting, organized, researched, supported and delivered effectively? Was CRISP-DM,
SEMMA, or a similar workflow followed to organize the work?
Grade Definitions
Students registered for undergraduate or graduate credit who complete the requirements
of a course may earn one of the following grades:
A Earned by work whose superior quality indicates a full mastery of the subject—and in
the case of A, work of extraordinary distinction. There is no grade of A+.
B+, B Earned by work that indicates a strong comprehension of the course material, a
good command of the skills needed to work with the course materials, and the student’s
full engagement with the course requirements and activities.
C+, C Earned by work that indicates an adequate and satisfactory comprehension of
the course material and the skills needed to work with the course materials, and that
indicates that the student has met the basic requirements for completing assigned work
and participating in class activities.
D+, D Earned by work that is unsatisfactory but that indicates some minimal command
of the course materials and some minimal participation in class activities that is worthy of
course credit.
E Earned by work that is unsatisfactory and unworthy of course credit. This grade may
also be assigned to students who do not submit required work in courses from which
they have not officially withdrawn by the withdrawal deadline. Zero or E grades are
assigned to students for missing work. These grades are included in the calculation of
the final grade.
It is the belief of the instructor that minus grades constitute a false precision in many
academic courses and further penalize frequent “A-“ students since there is no way to
obtain an “A+” to rebalance a GPA. To the student’s benefit, one can still earn a “plus” on
their final grade according to the scale above.
Course Materials
Machine Learning for Business Analytics
ISBN: 9781119835172
Authors: Galit Shmueli, Peter C. Bruce, Peter Gedeck, Inbal
Yahav, Nitin R. Patel
MACHINE LEARNING FOR BUSINESS ANALYTICS Machine
learning —also known as data mining or data analytics— is a
fundamental part of data science. It is used by organizations in
a wide variety of arenas to turn raw data into actionable
information. Machine Learning for Business Analytics:
Concepts, Techniques, and Applications in R provides a
comprehensive introduction and an overview of this
methodology. This best-selling textbook covers both statistical
and machine learning algorithms for prediction, classification,
visualization, dimension reduction, rule mining,
recommendations, clustering, text mining, experimentation, and
network analytics. Along with hands-on exercises and real-life
case studies, it also discusses managerial and ethical issues for
responsible use of machine learning techniques. This is the
second R edition of Machine Learning for Business Analytics.
This edition also includes: A new co-author, Peter Gedeck, who
brings over 20 years of experience in machine learning using R
An expanded chapter focused on discussion of deep learning
techniques A new chapter on experimental feedback techniques
including A/B testing, uplift modeling, and reinforcement
learning A new chapter on responsible data science Updates
and new material based on feedback from instructors teaching
MBA, Masters in Business Analytics and related programs,
undergraduate, diploma and executive courses, and from their
students A full chapter devoted to relevant case studies with
more than a dozen cases demonstrating applications for the
machine learning techniques End-of-chapter exercises that help
readers gauge and expand their comprehension and
competency of the material presented A companion website
with more than two dozen data sets, and instructor materials
including exercise solutions, slides, and case solutions This
textbook is an ideal resource for upper-level undergraduate and
graduate level courses in data science, predictive analytics, and
business analytics. It is also an excellent reference for analysts,
researchers, and data science practitioners working with
quantitative data in management, finance, marketing,
operations management, information systems, computer
science, and information technology.
Publisher: John Wiley & Sons
Publication Date: 2023-03-08
Edition: 2nd
THIS BOOK IS OPTIONAL.
This textbook has some overlap with lessons and should be
purchased by students wishing to expand beyond lessons to
add additional fluency and technical knowledge.
Academic Integrity Policy
You are responsible for understanding Harvard Extension School policies on Academic
Integrity and how to use sources responsibly. Violations of academic integrity are taken
very seriously. Visit Using Sources Effectively and Responsibly and the Harvard Guide to
Using Sources to review important information on academic citation rules.
AI Technologies. The Extension School's Academic Integrity Policy. prohibits students
from representing work as their own that they did not write, code, or create. It is never
permissible to submit work generated by machine learning and AI technologies (such as
ChatGPT) without proper attribution.
In our course we allow for the use and support of the BETA software, www.llmaid.com,
where students have the option of signing up (instructions given in class). This
technology uses chatGPT but is adjusted based on our course material. Thus chatbot
answers are returned along with references to specific lecture material and students are
encouraged to use this resource if desired.
Writing Code. While it may be common practice in non-academic settings to adapt code
examples found online or in texts, this is not the case in academia. In particular, you
should never copy code produced as coursework by other students, whether in the
current term or a previous term; nor may you provide work for other students to use.
Copying code from another student or any other source is a form of academic
dishonesty, as is deriving a program substantially from the work of another.
Writing code is similar to academic writing in that when you use or adapt code developed
by someone else as part of your assigned coursework, you must cite your source.
Paraphrasing without proper citation is just as dishonest with programming as it is with
prose. A program can be considered plagiarized even though no single line is identical to
any line of the source.
Accessibility Services Policy
The Division of Continuing Education (DCE) is committed to providing an accessible
academic community. The Accessibility Services Office (ASO) is responsible for
providing accommodations to students with disabilities. Students must request
accommodations or adjustments through the ASO. Instructors cannot grant
accommodation requests without prior ASO approval. It is imperative to be in touch with
the ASO as soon as possible to avoid delays in the provision of accommodation.
DCE takes student privacy seriously. Any medical documentation should be provided
directly to the ASO if a substantial accommodation is required. If you miss class due to a
short-term illness, notify your instructor and/or TA but do not include a doctor's note.
Course staff will not request, accept, or review doctor's notes or other medical
documentation. For more information, email [email protected].
Publishing or Distributing Course Materials Policy
Students may not post, publish, sell, or otherwise publicly distribute course materials
without the written permission of the course instructor. Such materials include, but are
not limited to, the following: lecture notes, lecture slides, video, or audio recordings,
assignments, problem sets, examinations, other students’ work, and answer keys.
Students who sell, post, publish, or distribute course materials without written permission,
whether for the purposes of soliciting answers or otherwise, may be subject to
disciplinary action, up to and including requirement to withdraw. Further, students may
not make video or audio recordings of class sessions for their own use without written
permission of the instructor.
Class Meeting Schedule
January 22: Introduction to R, R-studio & Git software.
We will set up our development environment on your local laptops & explain how each of
the software components are used. The class will execute some basic code examples.
January 29: Introduction to Data Mining & Exploratory Data Analysis (EDA)
We introduce more coding concepts including data types & functions. Then we start to
get summary statistics and build visualizations based on data.
February 5: More R Practice & EDA
The class will join data sets, and build more robust and interesting visualizations as part
of more sophisticated EDA.
February 12: Data Mining Workflows
As we increase our sophistication, before applying modeling and machine learning
methods in a business use case we learn the basic steps to prepare our data set
including dummy variables, feature engineering and imputation. By the end of class we
build our first simple machine learning model.
February 19: NO CLASS (University holiday)
February 26: Regression & Logistic Regression Modeling
The class will learn the basics of linear regression and how this traditional model can still
be useful for solving business problems. The class will follow code and execute the steps
of a model fit including future predictions after cleaning up realistic business data.
March 4: Decision Tree & Random Forest
Students will conceptually learn how a decision tree and random forest learn patterns
within a data set. The class will build and evaluate real machine learning or "AI" models
in a business context using all the code, data cleaning, EDA and machine learning
methods learned to date.
March 11: Spring Break - NO CLASS
March 18: Time Series Data
The class will learn about a new data class called time series which is useful in demand
forecasting. Code and explanations are given for predicting quarterly revenue and other
time series so students learn the basics of this robust field and how to identify when data
has a temporal element.
March 25: Equities
Students will use an API to obtain stock trading data and learn basic methods of
algorithmic trading. The goal is not to make day-traders of the participants but to
showcase how markets are highly efficient and technically supported despite retail
investors being naive to technical trading rules and methods.
April 1: Predicting Risk (Credit Default) & Non-Traditional Market
Simulation/Investing
The class covers building models to predict individual investment outcomes then
selecting optimal investments based on the relationship between risk & reward. Next, a
simple simulation is constructed to mimic a non-traditional investment market investing in
collectible trading cards so an investor can understand the outcomes before risking
capital.
April 8: Text Analysis & Natural Language Processing
As a new data set to the class, we explore how to extract insights and frequent phrases
from text. As a basic primer on natural language processing we discuss and
demonstrate code for "bag of words" methods.
April 15: Text Analysis & Natural Language Processing , Continued
We expand on the previous NLP lesson to include sentiment analysis & unsupervised
methods to identify topics using spherical K-Means and LDA topic modeling machine
learning methods. Time permitting we can introduce a document classification system
using elastic-net regression.
April 22: Data Sources, cURL & APIs
In this lesson the class will learn how to perform limited web-scraping, and obtain data
from unsecured and secured application program interfaces (APIs). Time permitting we
will demonstrate R performing a "cURL" request mimicking a web browser to obtain data.
April 29: Responsible AI & Technology Ethics
Technology ethics, and how to think about the responsible use of technology are
paramount in this AI driven world. As practitioners and consumers affected by AI, it is
important to have a grasp of the technology, its limitations and how to ethically apply
these technologies. This lesson covers some basic tenets of responsible AI and
students will purposefully build a sexist resume tool then apply methods to identify this
unjust (and illegal) model behavior.
May 6: Class Lab
The course material is very wide ranging yet dense. This sessions allows for flexibility to
cover topics that need additional attention, expand on use cases discussed, help with the
final case assignment. In the past students have come to class to ask about
approaching problems using the course techniques but applied to their own professional
process.
