Practice Problem 2 - Forecasting Ordroid Devices

3/15/2016
Practice Problem 2: Forecasting Ordroid Devices | Week 2 Practice Problems | CTL.SC1x Courseware | edX
MITx: CTL.SC1x Supply Chain Fundamentals
Week 2: Forecasting I - Introduction > Week 2 Practice Problems > Practice Problem 2:
Forecasting Ordroid Demand
Bookmarks
Bookmark
Course
Overview &
Logistics
Entrance
Survey
Week 1:
Overview of
Supply Chain
Management &
Logistics
Week 2:
Forecasting I Introduction
Welcome to Week
2
Lesson 1: Demand
Forecasting
Lesson 2: Time
Series Analysis
PRACTICEPROBLEM 2: FORECASTING ORDROID DEVICES

You have just been hired by a company that manufactures mid-range
communication devices that use the Ordroid open source operating
system. The company is focused on innovating its products and has not
put much thought on its inventory or forecasting capabilities.Your boss
thinks there might be a problem in the forecasting of the Ordroid Devices
and wants you to figure it out. The Ordroid, far from being new to the
market, has been out for two years.
Knowing this, you have asked for data on both years of historical sales as
well as any forecasts, promotions, pricing changes, or competitive
analyses made during this time. Your boss laughs and provides you with
all the data they have: the last six months of sales. You ask to meet with
the current demand planner for the Ordroid Devices and she tells you that
they use a forecasting algorithm of her own design and there is no
documentation.
Download the spreadsheets with the data here:
In Excel format (link toOrdroid_Data.xlsx)
In LibreOffice format (linkOrdroid_Data.ods)
Week 2 Practice
Problems
Supplemental
Materials for
MicroMasters
Part 1A
Week 2 Graded
Assignment
As a first step, you want to calculate some different performance metrics

for the small data sample. Recall that the definition of the error term is the
Actual demand minus the Forecasted demand.
Homework due Mar

02, 2016 at 15:00 UTC
Week 3:
Forecasting II -
What is the mean deviation of the forecasts in this data sample?
https://courses.edx.org/courses/course-v1:MITx+CTL.SC1x_2+1T2016/courseware/aeafb16ac1e64da2a16d98d6cdfa423b/b4a85bbcabc94dc6bf30fe713fb727a1/
1/7
3/15/2016
Exponential
Smoothing
Week 4:
Forecasting III Special Cases &
Extensions
Week 5:
Inventory
Management I
- Deterministic
Demand
Answer: 112.5
EXPLANATION
To find this, you just needed to find the error terms for each
observation. The average of the deviations (or error terms) is the
Mean Deviation.
You have used 3 of 3 submissions
Part 1B
What is the mean absolute deviation (MAD) of the forecasts in this data
sample?
Answer: 509.5
EXPLANATION
To find this, you just needed to average the absolute values of the
error terms for each observation.
Part 1C
What is the root mean square error (RMSE) of the forecasts in this data
sample?
Answer: 540.6
2/7
3/15/2016
EXPLANATION
To find this, you just needed to find average of the squared error
terms for each observation. The RMSE is simply the square root of the
average of the squared deviations (or error terms). We will use this
value extensively when we discuss inventory models.
Part 1D
What is the mean percent error (MPE) of the forecasts in this data sample?
Just enter the number and not the percentage sign, e.g., for 19.7% enter
19.7.
Answer: 4.1
EXPLANATION
To find this, you just needed to divide the error terms for each
observation by the actual demand for that period. The average of
these values is the MPE or mean percent error.
Part 1E
What is the mean absolute percent error (MAPE) of the forecasts in this
data sample? Just enter the number and not the percentage sign, e.g., for
19.7% enter 19.7.
Answer: 26.9
3/7
3/15/2016
EXPLANATION
To find this, you just needed to divide the absolute value of the error
terms for each observation by the actual demand for that period. The
average of these values is the MAPE or mean absolute percent error.
You can download the solution to this problem here:
In Excel format (link Ordroid_Solution.xlsx here)
In LibreOffice format (link Ordroid_Solution.ods here)
Part 2A
Based on your analysis, answer the following questions.
What can you say about the presence seasonality of demand?
There appears to be seasonality with lower demand in the
summer
There appears to be no seasonality
There is not enough data to evaluate for seasonality
EXPLANATION
There is not even one full year of data, so it would be very premature
to evaluate for seasonality! You need at least two full cycles to
determine seasonality.
4/7
3/15/2016
Part 2B
What can you say about the presence of a trend in the demand?
There appears to be a positive trend in demand
There appears to be a negative trend in demand

There does not appear to be any trend in demand
There is not enough data to evaluate for a trend
EXPLANATION
As opposed to seasonality, we do have enough data to suspect a

positive trend. In fact, if we compare
with
for the data, we see
that on average, the demand increases by about 176 units per month!
This averages to about 10% increase per month. So, yes, I would
suspect a positive trend here.
Part 2C
What can you say about the bias of the forecast?
The forecast appears to be positively biased
The forecast appears to be negatively biased
The forecast appears to not be biased
There is not enough data to evaluate for bias
5/7
3/15/2016
EXPLANATION
A bias is a persistent tendency to over or under predict. These

forecasts are not persistent in either. In fact, of the six periods, half
are over forecast and half are under forecast. So, there does not
appear to be any bias in the forecast.
Part 2D
What can you say about the accuracy of the forecast?
This appears to be a highly accurate forecast
This does not appear to be a highly accurate forecast

There is not enough data to evaluate for accuracy
EXPLANATION
This is not a very good forecast. The MAPE is almost 27% - pretty high.
But even worse, you can see that there is a pretty strong trend going
on and the forecasts totally ignore this.
However, since this is a relatively subjective question, we are giving
full points to all answers.

All Rights Reserved
6/7
3/15/2016
edX Inc. All rights reserved except where noted. EdX, Open edX and the edX and Open EdX logos are registered
trademarks or trademarks of edX Inc.
7/7

Practice Problem 2 - Forecasting Ordroid Devices

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Practice Problem 2 - Forecasting Ordroid Devices

Încărcat de

Drepturi de autor:

Formate disponibile

3/15/2016

MITx: CTL.SC1x Supply Chain Fundamentals

PRACTICEPROBLEM 2: FORECASTING ORDROID DEVICES

As a first step, you want to calculate some different performance metrics

Homework due Mar

What is the mean deviation of the forecasts in this data sample?

You have used 3 of 3 submissions

You have used 3 of 3 submissions

You have used 3 of 3 submissions

You have used 3 of 3 submissions

You have used 3 of 3 submissions

You have used 3 of 3 submissions

There appears to be a negative trend in demand

As opposed to seasonality, we do have enough data to suspect a

You have used 3 of 3 submissions

There is not enough data to evaluate for bias

A bias is a persistent tendency to over or under predict. These

You have used 3 of 3 submissions

This does not appear to be a highly accurate forecast

You have used 3 of 3 submissions

S-ar putea să vă placă și