You are on page 1of 34

Data Mining

and
Forecast Management

MGMT E-5070

Applied
Management
Science
for Decision
Making,
2012
Pearson
Prentice-Hall, Inc.
Applied Management
Science
for Decision
Making, 2e
2014 1e
Pearson
Learning
Solutions

Philip
A.A.Vaccaro
PhD
Philip
Vaccaro , ,PhD

What is Forecasting?

Forecasting
is the art
and science
of predicting
future events.

It may involve taking


historical data and
projecting it into the
future by means of a
mathematical model.
It may also be an
intuitive prediction.
It may also be a
mathematical model
adjusted by good
judgement.

Forecasting is Data Mining Too !


Data mining is the process of extracting
patterns or correlations among dozens
of fields in large relational data bases.
With the amount of data doubling every
three years, it is becoming increasingly
important for transforming data into information, which in turn, can be used to
increase revenues, cut costs, or both.
Data mining uses simple and multivariate linear, and non-linear regression
models as well as hypothesis testing.

Data Mining Example


one or more
Simple Linear Regression
models

A grocery chain analyzed local buying


patterns.
They discovered that when men bought
diapers on Thursdays and Saturdays,
they also tended to buy beer.

Data Mining Example


A
Multiple
Regression
Model

Further analysis showed that these men


usually did their weekly grocery shopping
on Saturdays. On Thursdays, however,
they only bought a few items.

Data Mining Example


The retailer concluded that the men purchased beer
to have it available for the upcoming weekend.
The grocery chain could use this newly discovered
information in various ways to increase revenue.
For example, they could move the beer display closer
to the diaper display. And, they could make sure that
beer and diapers were sold at full price on Thursdays!
INFORMATION to KNOWLEDGE to DECISION !

A Word of Advice

There is seldom a single superior forecasting


method. One firm may find exponential
smoothing to be effective. Another firm may
use several models, and a third firm may combine
both quantitative and subjective methods.
Whatever approach works best should be used.

Forecasting Time Horizons


1. Short-range forecast : Time span of up to 1 year but generally
less than 3 months. It is used for planning purchasing, job
scheduling, workforce levels, job assignments, and production levels.
2. Medium-range forecast : Time span of 3 months generally to
3 years. It is useful in sales planning, production planning /
budgeting, cash budgeting, and analysis of various operating
plans.
3. Long-range forecast : Generally 3 years or more in time span.
It is used in planning for new products, capital expenditures,
facility location or expansion, and research and development.

The Strategic Importance of Forecasting

Good forecasts are of critical importance in all


aspects of a business.

The forecast is the only estimate of demand


until actual demand becomes known.

Forecasts of demand therefore, drive the


decisions in many areas.

Forecast Impacts
Human Resources

Hiring, training, and terminating workers all depend


on anticipated demand. If the HR department must
hire additional workers without warning, the amount
of training declines and the quality of the workforce
suffers.

Forecast Impacts
Capacity

When capacity is inadequate,


the resulting shortages
can mean undependable
delivery, loss of customers,
and loss of market share.
When capacity is in excess,
costs can skyrocket.

Forecast Impact
Supply Chain Management
In the global marketplace, where expensive
parts for Boeing 787 jets are manufactured
in dozens of countries, coordination driven
by forecasts is critical.
Scheduling transportation to Seattle for final
assembly at the lowest possible cost means
no last-minute surprises that can harm already
low profit margins.

Product Life Cycle Influence

Products and even services, do not sell at a constant level


throughout their lives. Most successful products pass
through four stages : introduction, growth, maturity, and
decline.

Product Life Cycle Influence


Products in the first two stages
of the life cycle need longer
forecasts than those in the
maturity and decline stages.
Forecasts that reflect life cycle
are useful in projecting different
staffing levels, inventory levels,
and factory capacity as the
product passes from the first
to the last stage.

Forecasting Caveats
Forecasts are seldom perfect. Outside factors we
cannot predict or control often impact the forecast.
Most forecasting techniques assume that there is
some underlying stability in the system.
Product family and aggregated forecasts are more
accurate than individual product forecasts.
This approach helps balance the over and under
predictions of each.

Service Sector Forecasting


Barber Shops

Expect peak flows on Fridays and Saturdays.


Many call in extra help on the above days.
Most are closed on Sunday and Monday.

Service Sector Forecasting


Flower Shops

When Valentines Day falls on a weekend, flowers cannot be


delivered to offices, and customers are likely to celebrate
with outings rather than flowers ( low sales ) .
When Valentines Day falls on a Monday, some celebration
will have taken place on the weekend ( reduced sales ) .
When Valentines Day falls in midweek, busy midweek work
schedules make flowers the optimal way to celebrate
( higher sales ).

Service Sector Forecasting


Fast Food Restaurants

Use point-of-sale computers that track sales


every 15 minutes.
May use the moving average technique to
minimize the error of the 15-minute forecasts.
The forecasts are used to schedule staff, who
begin in 15-minute increments, not the 1-hour
blocks as in other industries.

Percent of Sales by Hour of Day

Hourly Sales at a Fast-Food Restaurant


20%
15%
10%
5%

11-12

1-2
12-1

( Lunchtime )

3-4
2-3

5-6
4-5

7-8
6-7

( Dinnertime )

9-10
8-9

10-11

Monday Calls at a FedEx Call Center

12%
11%
10%
9%
8%
7%
6%
5%
4%
3%
2%
1%
0%
1

A.M.

9 10

11 12 1

2 3 4 5 6

P.M.

9 10 11 12

Service Sector Forecasting


Federal Express

makes 1-year and 5-year models to predict number of service


calls, average handle time, and staffing needs.
breaks the forecasts into weekday, Saturday, and Sunday,
and then uses the Delphi Method and time-series analysis.
tactical forecasts are monthly, and use 8 years of historical
daily data. They predict caller volume by month, day of the
week, and day of the month.
the operational forecast uses a weighted moving average and
6 weeks of data to project the number of calls on a 30-minute
basis.

Service Sector Forecasting


Federal Express

Fed Exs forecasts are consistently


accurate to within 1% to 2% of actual
call volumes.

This means that coverage needs are


met, service levels are maintained, and
costs are controlled.

Forecasting Fundamentals & Models

TYPES
Time Series Models
Causal Models
Qualitative Models

Time-Series Models

Predict the future by using historical data.


Assume that what happens in the future is
a function of what has happened in the past.
Moving Average,
Weighted Moving Average,
Exponential Smoothing,
Trend Projection

Causal Models
Incorporate variables or factors that
might influence the quantity being
forecasted into the forecasting model.

The most common causal model is


regression analysis.
Ice cream sales, for example,
might depend on the season,
average temperature,
day of the week, and so on.

Qualitative Models
Incorporate judgmental or subjective factors
into the forecasting model.
Opinions by experts, individual experiences,
and other factors are expected to be very
important.
Used when accurate quantitative data are
difficult to obtain.
EXAMPLES ARE
THE DELPHI METHOD,
METHOD,
SALES FORCE COMPOSITE,
COMPOSITE,
AND
CONSUMER MARKET
SURVEY

1. The Delphi Method


Three types of participants: decision makers,
staff personnel, and respondents.
The decision makers make the actual forecast.
The staff personnel prepare, distribute, collect,
and summarize a series of questionnaires and
survey results.
The respondents are those whose judgements
and values are being sought. They provide input to the decision makers before the forecast
is made.

The
Delphi Method

2. Sales Force Composite


Each salesperson estimates what sales
will be in his or her region.
These forecasts are reviewed to ensure
that they are realistic.
These forecasts are combined at the
district and national levels to reach an
overall forecast.

3. Consumer Market Survey

This method solicits input from customers


or potential customers regarding their future
purchasing plans.
It can help not only in preparing a forecast
but also in improving product design and
planning for new products.

Consumer
Market
Survey

Consumer
Market
Survey

Types of Forecast Models


Causal Methods

Regression
Analysis
Multiple
Regression

Time Series
Methods
nave approach
arithmetic mean
moving average
weighted average
weighted - moving
average
exponential smoothing
trend projections
decomposition

Qualitative
Models
delphi method
jury of executive
opinion
sales force
composite
consumer market
survey

Cost vs. Accuracy Tradeoff


HIGH

TOTAL
of
MODEL COST
and
FORECAST
ERROR COST

ECONOMETRIC
MODELS

M
O
D
E
L

OPERATING COSTS
DUE TO INACCURATE
FORECASTS

CAUSAL
MODELS

SOPHISTICATED
TIME
SERIES

C
O
S
T
S

SIMPLE
TIME
SERIES

THE
OPTIMAL
REGION

QUALITATIVE
MODELS

LOW
100%

DECLINING ACCURACY

0%

You might also like