Scribd
Upload
186 views4 pages

Forecasting Notes

(1) Forecasting involves predicting future demand based on historical data patterns, though forecasts are rarely fully accurate. Short-term and group forecasts tend to be more accurate. (2) There are qualitative and quantitative forecasting methods. Quantitative methods analyze previous numerical data using techniques like time series analysis and causal/regression models. (3) Time series forecasts can be created using simple techniques like the naive method or moving averages that smooth data using a weighted formula. Causal methods like regression analyze correlation between variables.

Uploaded by

krithika1288
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
186 views4 pages

Forecasting Notes

(1) Forecasting involves predicting future demand based on historical data patterns, though forecasts are rarely fully accurate. Short-term and group forecasts tend to be more accurate. (2) There are qualitative and quantitative forecasting methods. Quantitative methods analyze previous numerical data using techniques like time series analysis and causal/regression models. (3) Time series forecasts can be created using simple techniques like the naive method or moving averages that smooth data using a weighted formula. Causal methods like regression analyze correlation between variables.

Uploaded by

krithika1288
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd

Forecasting:

(1) Predict the future demand


(2) Based on previous data (historical data)
(3) Never fully accurate (Rarely accurate)
(4) Forecasting for short-term is more accurate
(5) Forecasting for group of items / whole unit is more accurate compared to individual
departments

Ways:

(1) Qualitative – Expert opinions, market research, Delphi method


(2) Quantitative method – previous data, objective in nature

Types:

(1) Short-term – 3 to 6 moths


(2) Medium –term – 1 -2 years
(3) Long-range – 5 to 10 years

Quantitative Techniques:

(1) Time series


(2) Causal methods / associative methods

Time series:

(1) Naïve method


(2) Simple moving average
(3) Weighted moving average
(4) Exponential moving average / Exponential smoothing

Causal method:

(1) Regression

Naïve:

Future demand = Current actual demand

Forecast Data:

Data = Historic data + random variation

Historic data / time series can follow patterns such,

(1) Horizontal – very less fluctuation ; mostly the data has the same average
(2) Trend
(3) Seasonal
(4) Cyclic
(5) Random

Naïve: - usage of past data as future forecast

Horizontal – naïve
Trend:

Moving average

Average / Mean:

A = sum of all the values / number of items

e.g. 5,6,7

5+6+7/3 = 6

5,5,5

A=5

3,4,2  2,3,4

A=3

1,2,1  1,1,2

A=1

Age of three children in a play group

3,2,20

A=5

Exponential smoothing

Works based on,

(1) Last year’s forecast


(2) Last year’s actual demand
(3) Alpha / smoothing constant ( set between 0 and 1)

Current forecast = last yr forecast + alpha X (Last yr actual demand – last yr forecast)

Difference between actual demand and the forecast—is called? -------- Variation --- Error!

Causal method:

Linear Regression:

Variable? – Something that keeps varying / changing (Marikite irukum!)

Assume there are two variables,

V1, V2

IF the variation in V1 has some connection to the variation in V2 it is possible that V1 and V2 are
associated with each other

When two numerical variables (number based variables) are associated with each other we call it
correlation!
IQ and Age – correlation between age and IQ

WT and BMI – as weight increases the BMI also increases

Regression:

It will draw a line based on the data

Positive correlation:

IQ
160
150
140
130
120
110
100 100

80

60

40

20

0
0 5 10 15 20 25 30 35

Increase in V1 causes increase in V2

Negative correlation

Health quality
120
100
100
80
80
60
60
40
40

20

0
0 5 10 15 20 25 30 35

Increase in V1 decreases v2

Un correlated
V2
1.2

1 1 1 1 1

0.8

0.6

0.4

0.2

0
0.5 1 1.5 2 2.5 3 3.5 4 4.5

Changes in V1 doesn’t cause changes in V2

You might also like