Académique Documents
Professionnel Documents
Culture Documents
Tahir Hameed
Nirbhay Pratap Singh
Pallavi Sharma
Pratyush Thakur
What Is Forecasting?
business
decisions
• Basis for most
important
planning
What is forecasting?
Sales will
Could be a be $200
Could be a
prediction Million!
prediction
based on past based on
behaviour and expertise and
mathematical intuition
models
Consumer
Market Survey
Quantitative Forecasting
methods
Quantitative Forecasting Methods can be classified in to two
categories
1. Time series analysis -
Analysis of the past behavior of the data over time and
forecasting under the assumption that the forces that
generated the past data will also generate the future -
Analysis of past demand pattern provides a good basis for
forecasting future demand
2.
Causal models:
Associate models – Regression – Simple and Multiple
Components of Forecasting
Demand
1.Time Frame
• Short to mid range forecast
-eg. HP- 12 to 18 months
2. Demand Behaviour
Demand Behaviour
Trend Cyclical
Seasonal Random
Demand Behaviour- Trend
Quantity
Time
after year.
Year 1
Quantity
Year 2
| | | | | | | | | | | |
J F M A M J J A S O N D
Months
Seasonal: Data consistently show peaks and valleys at
the same time each year.
Cyclical Variation
Repetitive fluctuations or swings of varying length
Quantity
| | | | | |
1 2 3 4 5 6
Years
•
Quantity Random or Chance component
Time
Forecasting Process
1. Identify the 2. Collect historical 3. Plot data and
purpose of forecast data identify patterns
Yes
9. Adjust forecast 10. Monitor results
8a. Forecast over and measure
planning horizon based on additional
qualitative forecast accuracy
information and
insight
Time Series Analysis
Forecasting Models –
1. Moving averages
Disadvantage:
Ft +1 = α Dt + (1 - α )Ft
where:
Ft +1 = forecast for next period
Dt = actual demand for present period
Ft = previously determined forecast for
present period
α = weighting factor, smoothing
constant
Linear Trend Line
• Linear regression is a method of forecasting in
which a mathematical relationship is
developed between demand & factors
causing demand behavior.
• If a time series exhibits a linear trend, the
method of least squares may be used to
determine a trend line (projection) for future
forecasts.
•
Seasonal adjustments
• Seasonal pattern is a repetitive
increase and decrease in demand.
(monthly, weekly, daily, weather,
festival etc)
– Sales of apparels
– Sales of greeting cards
–
• The seasonal fluctuations are
adjusted by multiplying the Di normal
forecast
S e a so nby a cto
a lfa seasonal
r = S = factor.
i D
Forecast Accuracy
• Forecast error
Σ | Dt -
– difference between forecast and actual demand Ft |
MAD =
– MAD n
• mean absolute deviation (avg absolute difference
between the forecast & demand)
|Dt - Ft|
• Smaller the MAD value relative to magnitude of the
data more accurate is the forecast
– MAPD MAPD =
• mean absolute percent deviation(absolute error as aDt
%age of demand rather than period)
• Eliminates the problem of interpreting the measure of
accuracy relative to the magnitude of demand &
forecast values.
E = et
• Lower %age deviation implies more accurate forecast.
– Cumulative error
• Summing the forecast errors, large ±E indicates
forecast is highly biased.
et
• Large +E shows biased low/Large –E shows biased
high. E = n
– Average error or bias
• Averaging cumulative error over no of periods
Forecast Control
• Tra ckin g sig n a l
n m o n ito rs th e fo re ca st to se e if it is
b ia se d h ig h o r lo w
n ( Dt - Ft) E
Tra
n
ckin g sig n a l = MAD = MAD
n R e co m p u te d e a ch p e rio d
n 1 M A D ≈ 0 .8
m o st fre q u e n tly
Statistical Control Charts
• ( Dt - Ft)2
• = n - 1
•
•
Using we can calculate statistical
control limits for the forecast
error
Control limits are typically set at
3
Regression Methods
Linear regression
a mathematical technique that relates a dependent variable to an
independent variable in the form of a linear equation
y = a + bx
Multiple Regression
Study the relationship of demand to two or more independent
variables
y = β 0 + β 1x1 + β 2x2 … + β kxk
Where
β 0 = the intercept
β 1, … , β k = parameters for the independent variables
x1, … , xk = independent variables
Correlation and Coefficient
of Determination
Correlation, r
Measure of strength of relationship
Varies between -1.00 and +1.00
Coefficient of determination, r2
Percentage of variation in dependent
variable resulting from changes in
the independent variable
Multiple Regression
Study the relationship of demand to
two or more independent variables
β 0 = the intercept
β
1, … , β k = parameters for the
independent variables
x1, … , xk = independent variables
•