Professional Documents
Culture Documents
Forecasting: Dr. Chinmaya Kumar Sahoo Department of Mechanical Engineering NIT Silchar
Forecasting: Dr. Chinmaya Kumar Sahoo Department of Mechanical Engineering NIT Silchar
This class notes are only for academic propose and class room circulation
Thi s Photo by Unknown Author i s licensed under CC BY-SA-NC
Forecasting
• It indicate prediction of future event.
• The prediction is one of the first decision making activity used for
production system, material requirement planning, scheduling etc.
Types
• Technology forecast
• Economic forecast: Mostly done by Govt. agencies to predict the
economic activity.
• Demand forecast: Used to predict the expected level of demand for
goods or services
Sources of data
• Forecast is done based on the past data.
• Data could be obtained form the old records, survey journals, Govt.
publication, news paper etc.
Forecasting Models
• Qualitative technique: Where no data are available. E.g. Delphi
method, market survey
Forecast variable
Time
3. Cyclical pattern:
• In this pattern, length of cycle is more than one year. The cycle does
not repeat at constant interval of time.
e.g. metals, gross national product.
Forecast variable
Time
• 4. Trend pattern
The pattern shows when there is increase or decrease of variable over
time.
e.g. sales of many products, stocks prices, business etc.
Forecast variable
Time
Forecasting Error
• Error = 𝑒𝑡 = 𝐷𝑡 − 𝐹𝑡
𝑒𝑡 = forecast error for time t
𝐷𝑡 = Demand for the period t
𝐹𝑡 =Forecast for period t
• Mean Absolute Deviation (MAD)
1 𝑛 𝐷𝑡 −𝐹𝑡
𝑀𝐴𝑃𝐸 = σ𝑖=1 × 100
𝑛 𝐷𝑡
Simple moving average method
• This method considers the most recent data in series are considered.
1
• 𝑀𝑡 = 𝐷𝑡−(𝑛−1) + 𝐷𝑡−(𝑛−2) + 𝐷𝑡−(𝑛−3) … … … . 𝐷𝑡
𝑛
• Ft+1 =
• Calculate a weighted average forecast for the data using a weight of
0.6 for most recent data and weight of 0.3 and 0.1 for successive
older data.
• Calculate the forecast value for the 4th week
• Solution:
F3= 500, D3= 480, 𝛼 = 0.2
F4= F3 + 𝛼 (D3- F3) =
= 496
Question
• A shopkeeper found that Demand and forecast value for the month of
November is 890 and 876 respectively. Find the forecast value for the
month of December.
Use the smoothing constant value as 0.2
Year 1 2 3 4 5 6 7 8 9 10
Act. 10 12 8 11 9 10 15 14 16 15
demand
(Lakh)
Regression Analysis (Least Square method)
• Forecast demand could be represented by the equation
• 𝐹𝑡 = 𝑎 + 𝑏. 𝑡; Y = a + b. x
x = independent variable
y = dependent variable
a = intercept
b = slop (trend)
σ 𝑥.𝑦 −𝑛.𝑥.
ഥ 𝑦ത
• 𝑎 = 𝑦ത − 𝑏. 𝑥ҧ , 𝑏 = σ 𝑥 2 −𝑛.𝑥ҧ 2
,
• A company found that annual profit depends on the expenditures for
research. Information for the least six years is given below. Estimate
the profit when expenditure on research is 6 units.
Year Expenditure for Research Annual Profit (y)
(x)
1 2 20
2 4 30
3 5 40
4 10 60
5 9 55
6 7 43
7 8 ?
Year Annual x.y 𝒙𝟐
Expendit Profit 37
ure for (y) 𝑥ҧ = = 6.16
6
Researc 248
h (x) 𝑦ത = = 41.3
6
ഥ𝑦
σ 𝑥.𝑦 −𝑛.𝑥. ത
1 2 20 40 4 𝑏=
σ 𝑥2 −𝑛.𝑥ҧ 2
= 4.808
2 4 30 120 16
3 5 40 200 25 𝑎 = 𝑦ത − 𝑏. 𝑥ҧ = 11.682
4 10 60 600 100
5 9 55 495 81
6 7 43 301 49 When expenditure is 6 unit
Total 37 248 1754 275 y = a + b.x = 11.68 + (4.808 x 6) = 40.52
Inventory Management
• Inventory management is under the material management.
• The inventory includes Raw material, semi finish materials, products,
packaging etc.
• The materials cost is approximately 60% of the production cost. So
proper management is required.
Type of inventory
• Raw material inventory
• Ordering Cost