You are on page 1of 2

ASSIGNMENT TEST-1

Time: 1 hour Marks: 20

1. Describe the Delphi method of forecasting. [2]


2. Khethan Industries has demand data of a product and is given below. This firm planned to
use four-month weighted moving average and exponentially weighted moving average for
forecasting the demand and decided to evaluate these forecasts based on MSE. Forecasted
demand estimated for the months, May to October, is planned to be used for the evaluation.
Which is the best forecast method for the given data based on the evaluation? Determine
the forecast for November and December using the best method identified in the evaluation.
Also give a three-sigma interval estimate of the forecast of demand of these months when
the best method identified is used. Use MAD for estimating standard deviation. [8]
Month Actual Demand
January 175
February 165
March 190
April 210
May 205
June 230
July 220
August 225
September 215
October 240
In the case of the four-month weighted moving average, the weights for the data are
0.15, 0.25, 0.3 & 0.3. Most recent demand has higher weight than distant past data.
In the case of the exponentially weighted moving average, α = 0.2, and the initial
forecast for May month is the simple arithmetic average of the past four months. The
manager of the firm is interested to know the following also:
a) What are the weights for the past demand data in the exponentially weighted
moving average estimated for October? [2]
b) What will be the value of α if six-month moving average is found to be giving a
good estimate for demand but wants to use exponentially weighted moving average
instead of the six-month moving average? [1]
c) If the above demand data shows a trend and randomness, then what is the equation
of the demand process? [1]
d) Plot the data and the firm is interested to have your comment on the decision of
using weighted moving average or exponentially weighted moving average. [2]
3. The registrar at State University believes that decreases in the number of fresh applications
that have been experienced are directly related to tuition fee increases. They have collected
the following application and tuition fee data for the last 10 years:
Year No. of Applications Annual Tuition Fee (in INR)

1 6010 36000
2 5560 36000
3 6100 40000
4 5330 44000
5 4980 45000
6 5870 57000
7 5120 60000
8 4750 60000
9 4615 75000
10 4100 80000
Develop a linear regression model for these data and forecast the number of applications
for State University if tuition fee increases to 90,000 per year and if tuition fee is
lowered to 70,000 per year. [4]

You might also like