Overview
The late 1900s heralded in an era of tools for
making better-informed decisions in business.
Commercial products became available under
the umbrella terms of decision support systems
(DSS), group decision support systems (GDSS),
executive information systems (EIS) and expert
HOME STUDENT RESOURCES HELP STUDENT HELP
ANUSHA CLOASGUOLUAT 10
systems (ES). Most commercial ES were rulebased
systems, whereby both the knowledge
and the problem-solving procedures were
stored in the form of rules. ES provided
explanatory capabilities, which were essential to
the transfer of expertise and problem solving
(Curry, 1993). However, dramatic advances in
the !elds of machine learning, statistical
modelling and data mining (propelled by
advances in storage technology) have all but
replaced last century’s tools.
Today, experts in data science, arti!cial
intelligence (AI) and big data are among the
most sought-after roles in organisations from all
industry sectors as well as government
(Davenport & Patil, 2012). In line with Laureate’s
mission of giving our students an edge in
employability, this module discusses practical
applications of data-driven decision tools (Das,
2014) and integrates concepts from previous
modules to help understand where these tools
!t into decision-making.
Data science, AI and big data have di”erent
roots in their own distinct data-driven traditions.
Arrttii!cciiall IInttelllliigencce is enabling computers to
think. For example, the aforementioned (and
somewhat obsolete) rules-based Expert Systems
form part of AI. Macchiine LLearrniing is a subset
of AI, which uses statistical tools to learn from
data. In 2006, Deep LLearrniing has emerged as
a part of machine learning, which employs
multi-layer neural networks. This is the area
where most advances have been made in AI. For
example, when Google’s AI is commonly
example, when Google’s AI is commonly
referred to in the literature, the reference is
mostly to Google’s deep learning capabilities.
Datta Scciiencce is a di”erent discipline from AI
although it has some overlap with AI and
machine learning (less so with deep learning). It
follows an interdisciplinary approach, which lies
at the intersection of maths, statistics, software
engineering and systems thinking. Data Science
deals with data collection, data cleaning,
analysis, visualisation, model creation, model
validation, prediction, designing experiments,
hypothesis testing and much more. The aim of
all these steps is to derive insights from data.
However, the amount of data collected by
companies today is so vast it creates a large set
of challenges regarding data acquisition,
storage, analysis and visualisation. Biig Datta
involves four ‘V’s – Volume, Variety, Veracity and
Velocity – which requires it to be treated
di”erently from conventional data as follows:
Volume: The amount of data involved
here is so huge, that it requires
specialised infrastructure to acquire,
store and analyse it. Distributed and
parallel computing methods are
employed to handle this volume of data.
Variety: Data comes in various formats;
structured or unstructured. Structured
means neatly arranged rows and columns
in so-called relational databases.
Unstructured means that it comes in the
form of raw text, videos and images,
which requires di”erent database
systems than relational ones.
systems than relational ones.
Veracity: Data needs to be correct for it to
be meaningful. The industry term for this
is “garbage in [bad data], garbage out
[wrong conclusions]”. Care needs to be
taken to make sure the data captured is
accurate and data integrity is maintained
— especially, as the data increases in
volume and variety.
Velocity: According to IBM (2016), 90% of
data in the world was created in the
previous two years alone (Loechner,
2016). This velocity of information
generated is bringing its own set of
challenges. For some businesses, realtime
analysis is crucial. Any delay will
reduce the value of the data and its
analysis for business.
The jargon of commercial decision-making tools
in the marketplace can be positioned on
descriptive, diagnostic, predictive or prescriptive
dimensions as depicted in Figure 1 below.
FFiig 1:: Datta–drriivven ttoollss fforr decciissiion
makkiing
(Kaduk, 2016)
These product concepts all aspire to converge
towards the top of the DIKW pyramid (refer
Module 1) in a multidisciplinary approach as
shown in Figure 2 below:
FFiig 2:: Convverrgencce off ccommerrcciiall ttoollss
ttowarrdss tthe ttop off tthe DIIKW pyyrramiid
(Press, 2016)
Such technologies are clearly reinvigorating a
rational decision-making model, which may
leave us vulnerable to algorithmic & data biases
as well as overcon!dence vis-à-vis blind spots
(refer Module 2). Consider this: human choices
and biases a”ect errors already in the
technology factory. This causes Type I (false
positives) and Type II errors (false negatives or
opportunity cost), which we discuss in this
module. Therefore, we reinforce the importance
of intuition and balanced decision-making with
data-driven tools. Blind trust in (or outright
rejection of) data-based technology is a
consequence of several myths (Kelleher &
Tierney, 2018):
We can let data tools run over data
We can let data tools run over data
without skilled human oversight at
several stages of the process;
Data science = big data = arti!cial
intelligence;
Data tools are easy to use & do not
require deep domain expertise;
Data tools pay for themselves quickly
even in the absence of a well-understood
business problem and appropriate data.
The last point above about the creation of databased
technology takes us back to groupdecisions
from Module 3. Consider “group think”
or a lack of diversity in the data science factory,
for example. This causes a statistical error,
which is then compounded by a decisionmaker’s
own judgement, interpretation and
intuition bias when decisions are made from the
tool’s !ndings.
When data-driven technologies are !nally rolled
out into the workplace, several change
management challenges arise as depicted in
Figure 3 below (also refer Soft Systems
Methodology in MGT603: Systems Thinking):
FFiig 3:: Change managementt & grroupss iin AII
ttrranssfforrmattiion
(Fountaine, McCarthy, & Saleh, 2019)
For ethical issues involved with data-driven tools
recall the learnings from Module 4. Supervised
learning-based data tools are perpetuating
“what has been”, meaning the data is leaning
towards keeping the status quo. But what if the
status quo is not the desirable state? With AI we
learn from past data and then deduce the future
from the past – and thus AI keeps things the way
they have been. With unsupervised learningbased
data tools, on the other hand, these tools
keep learning (and make their own decisions
from new data) after they have left the data tool
factory. Data Science resolves this through
human interaction in problem solving in a
consultative process, but AI does not go as far to
the same degree (Chowdhury & Mulani, 2018):
Data is not objective, is it re#ective of preexisting
social and cultural biases?
AI can be a method of perpetuating bias,
leading to unintended negative
consequences and inequitable outcomes.
The !eld of AI ethics draws an
interdisciplinary group of lawyers,
philosophers, social scientists,
programmers, and others. Figure 4 below
depicts why this is required, considering
the simple data-based decisions made by
AI.
FFiig 4:: Whatt AII decciidess wiitthoutt rregarrd tto
etthiiccss
(Ng, 2016)
References
Chowdhury, R., & Mulani, N. (2018, October 24).
Auditing algorithms for bias. Harvard
Business Review. Retrieved from
https://hbr.org/2018/10/auditingalgorithms-
for-bias
Curry, D. J. (1993). The new marketing research
systems: How to use strategic database
information for better marketing
decisions. Somerset, NJ: John Wiley &
Sons
Das, S. (2014). Computational business analytics.
Boca Raton, FL: CRC Press
Davenport, T., & Patil, D. J. (2012, October). Data
Scientist: The sexiest job of the 21st
century. Harvard Business Review, 90
(10), 70–76
Fountaine, T., McCarthy, B., & Saleh, T. (2019,
July/ August). Building the AI-powered
organization. Harvard Business Review,
62–73
Kaduk, T. (2016, December 14). 4 stages of Data
Analytics maturity: Challenging
Gartner’s model. Retrieved from
Linkedin:
https://www.linkedin.com/pulse/4-
stages-data-analytics-maturitychallenging-
gartners-taras-kaduk/?
trackingId=OB0BeGS2rIMMauGR56UIzA%3D%3D
Kelleher, J. D., & Tierney, B. (2018). Data Science.
Cambridge, MA: The MIT Press
Loechner, J. (2016, December 22). 90% of today’s
data created in two years. Research
Brief. Centre for media Research.
Retrieved from
https://www.mediapost.com/publications/article/291358/90-
of-todays-data-created-in-twoyears.
html
Ng, A. (2016, November 9). What arti!cial
intelligence can and can’t do right now.
Harvard Business Review. Retrieved
from https://hbr.org/2016/11/whatarti!
cial-intelligence-can-and-cant-doright-
now
Press, G. (2016, October 17). Visually linking AI,
Machine Learning, Deep Learning, Big
Data and Data Science. What’s the Big
Data. Retrieved from
https://whatsthebigdata.com/2016/10/17/visuallylinking-
ai-machine-learning-deeplearning-
big-data-and-data-science/
Essential Resources:
Essential Resources:
Describes decision-making tools
in the late 1900’s before
Business Intelligence, Arti!cial
Intelligence and Data Science emerged.
Curry, D. J. (1993). The new marketing
research systems: how to use strategic
database information for better marketing
decisions. Somerset, NJ: John Wiley &
Sons.
This article explains the
employability opportunities of
data-based degrees Davenport,
T. H., & Patil, D. J. (2012). Data Scientist:
The sexiest job of the 21st century.
Harvard Business Review, 90 (10), 70–76
This book introduces data
science, explaining its evolution,
relation to machine learning,
current uses, data infrastructure issues,
and ethical challenges.
Kelleher, J., & Tierney, B. (2018). Data
science (Essential knowledge series).
Cambridge, MA: MIT Press.
Optional Resources
This article provides a market
overview, examines how much
overview, examines how much
data is involved, how much
might be useful, what tools and
techniques are available to analyse it, and
whether businesses are getting to grips
with big data. McLellan, C. (2013). Big data:
An overview. Retrieved from ZDNet:
https://www.zdnet.com/article/big-dataan-
overview/
This clip explains what a data
scientist does.
Simplilearn. (2018). Data science
in 5 minutes. Retrieved from YouTube:
https://youtu.be/X3paOmcrTjQ
This clip explains the di”erence
between AI and Data Science.
Applied AI Course. (2017).
Arti!cial Intelligence Vs Machine Learning
Vs Data science Vs Deep learning.
Retrieved from YouTube:
https://youtu.be/QizsAE4fBpQ
This clip explains what Big Data
is.
World Economic Forum. (2016).
What is big data? Retrieved from YouTube:
https://youtu.be/eVSfJhssXUA
Supervised versus unsupervised
Supervised versus unsupervised
learning in arti!cial intelligence.
The clip explains data-induced
bias issues in supervised learning and why
that method perpetuates the status quo –
even if is not the desirable state.
Unsupervised learning causes a danger/
risk of continuously learning from new
data, which may cause ethical issues.
Big Data University. (2017). Machine
learning – supervised vs unsupervised
learning. Retrieved from YouTube:
https://www.youtube.com/watch?
v=cfj6yaYE86U&feature=youtu.be
Deep learning with multilayer
neural networks. The clip
explains how image and natural
speech recognition is done, plus how
neural networks work. It’s all the rage in
the industry today and Google, Alexa and
Siri are leading the market.
3Blue1BROWN. (2017). But what is a
Neural Network?. Retrieved from
YouTube:
https://www.youtube.com/watch?
v=aircAruvnKk&feature=youtu.be
Learning Activities:
Learning activities are not part of summative/
graded assessment; however, they are designed
to prepare you for incremental graded
assessment and expand your learning. These
assessment and expand your learning. These
activities encourage a community learning
experience between peers and provide
opportunities for facilitators to o”er formative
feedback to the student cohort. Students should
spend 1-2 hours per module on learning
activities.
LLearrniing Exxerrcciisse 6..1:: Datta
Scciiencce,, Biig Datta,, AII iin yyourr
worrkkpllacce
Review the resources above.
- Does your organisation use any
tools from Data Science, Big
Data or Arti!cial Intelligence –
or is it all Microsoft Excelbased? - What are the advantages of
these rational, data-based
decision-making tools in your
organisation? - What are the dangers of relying
only on such tools? - What part does intuition play in
wise decision-making,
supported by the evidence
from data-based decisionmaking
tools?
Record your responses in Discussion
Forum 6.1
To participate in the Discussion
Forum, click here to scroll to the
bottom of this page then click on the
bottom of this page then click on the
“Moodduullee 66 — Diissccuussssiioonn FFoorruum” link.
LLearrniing Accttiivviittyy 6..2:: Datta
Scciiencce
Review the resources above - What sort of dangers do you
see with data-based decision
tools? - What consultations with
employees have taken place in
relation to what data are
collected, how they are stored,
accessed, and their application
in your workplace? - How do bias and group
decision-making impact
conclusions made with databased
decision tools? - What ethical and security
considerations are there in
relation to big data in your
workplace? - Can you give an example for an
opportunity cost (type II error)
arising from a wrong
conclusion at your work or in
your life? This does not have to
do with the tools we discussed. - Where could you go to !nd out
more on Data Science, Big Data
and Arti!cial Intelligence? - How about resources for
learning about emerging career
opportunities?
Record your response in Discussion
Forum 6.2
To participate in the Discussion
Forum, click here to scroll to the
bottom of this page then click on the
“Moodduullee 66 — Diissccuussssiioonn FFoorruum” link.
The post Data Science and Artificial Intelligence in Organisational appeared first on My Assignment Online.
