Professional Documents
Culture Documents
Business
Dr. Anandakuttan B Unnithan
Indian Institute of Management
Kozhikode
Business and Numbers
“when you can measure what you are speaking
about, and express it in numbers, you know
something about it; but when you cannot measure it,
when you cannot express it in numbers, your
knowledge is of a meagre and unsatisfactory kind; it
may be the beginning of knowledge, but you have
scarcely in your thoughts advanced to the state of
Science, whatever the matter may be."
Business
Analytics
Hypothesis-Based vs. Exploratory-
Based
• The hypothesis-based method:
– Formulate a hypothesis of interest.
– Design a research study or experiment that will yield data
to test this hypothesis.
– Accept or reject hypothesis depending on the outcome.
– Testing Models on the data; Models represent theories
• Exploratory-based method:
– Try to make sense of a bunch of data without an a priori
hypothesis!
– The only prevention against false results is significance:
• ensure statistical significance (using train and test etc.)
• ensure domain significance (i.e., make sure that the results
make sense to a domain expert)
Hypothesis-Based vs. Exploratory-
Based
• Experimental Scientist:
– Assign level of fertilizer randomly to plot of land.
– Control for: quality of soil, amount of sunlight,...
– Compare mean yield of fertilized and unfertilized plots.
• Data Miner:
– Notices that the yield is somewhat higher under trees
where birds roost.
– Conclusion: droppings increase yield.
– Alternative conclusion: moderate amount of shade
increases yield.(“Identification Problem”)
What is Decision Modeling?
A scientific approach to managerial decision
making
Alerts
What actions are
needed?
Query/drill down
Where exactly is the
problem?
Ad hoc reports
How many, how often, where?
High Low
Performers Performers
65% Have significant decision-
support/analytical capabilities 23%
36% Value analytical insights to a very large extent
8%
77% Have above average analytical capability within industry
33%
77% Have BI/Data Warehouse modules installed 62%
73% Make decisions based on ES data and analysis
51%’
40% Use analytics across their entire organization
* Based on an Accenture study with a sample of
23%
400 companies worldwide (2005). Source: Tom
Davenport.
The Connected Analytics
Process
Profit
Effectiveness
Action Improve Competitive advantage
guidelines/Reports Company
What if capabilities Performance New revenue
Integration with Cost
company processes reduction
Productivity
Support gains
Guide
Opportunity
Implementation Opportunities Identification
for
Analytics
Structured
process Models
Real-time Promote Build an Data
Interactive Reasoned Analytics Digital
Distributed Decisions Foundation
across the infrastruct
organization ure
Analytics
Business Environment
Automatic scanning, data entry,
subjective interpretation
Business
Analytics Data Database management, e.g..,
selection, sorting, summarization,
report generation
Information
Decision model; mental model
Fixed
Population Size
Innovat Imitat
ors Innovat ors
ors
Influenc
Timing of e Timing of Purchases
Purchases by Imitator by
Innovators s Imitators
Pattern of Sales
Growth
of New Product
Graphical Model
Cumulative Sales
of a Product
Fixed
Population Size
Time
Mathematical Model
dxt
= (a + bxt)(N – xt)
dt
N = Population size
Max
Sales
Response
Respo
nse
Current Functi
Sales on
Min
Curre
nt
Effort
Effort Level
A Simple Model
Y (Sales Level)
} b (slope of
a
}
the
(sales level when 1 salesline)
advertising = 0)
X (Advertising)
Phenomena
P1: Through Origin P2: Linear
Y Y
X X
P3:Decreasing Returns
P4: Saturation
(concave)
—
Q
Y Y
X X
Phenomena
P5:Increasing Returns
P6: S-shape
(convex)
Y Y
X X
Y Y
X X
Linear Model
Y = a + bX
• Linear/through origin
Fractional Root Model
Y = a + bXc
Y = aebx; x > 0
Y = a (1 – e–bx) + c
Xc
d + Xc
Y = b + (a–b)
Widely used.
Y = af(X1) + bg(X2)
Y = aX 1b X 2c
b and c are elasticities.