practice problem 2_ forecasting ordroid devices

8/16/2019 Practice Problem 2_ Forecasting Ordroid Devices

1/7

3/15/2016 Practice Problem 2: Forecasting Ordroid Devices | Week 2 Practice Problems | CTL.SC1x Courseware | edX

https://courses.edx.org/courses/course-v1:MITx+CTL.SC1x_2+1T2016/courseware/aeafb16ac1e64da2a16d98d6cdfa423b/b4a85bbcabc94dc6bf30fe713fb727a1/

Bookmarks

Week 2: Forecasting I - Introduction > Week 2 Practice Problems > Practice Problem 2:

Forecasting Ordroid Demand

Bookmark

PRACTICE PROBLEM 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 to Ordroid_Data.xlsx)

In LibreOffice format (link Ordroid_Data.ods)

Part 1A

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.

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

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

Week 2 Practice

Problems

Supplemental

Materials for

MicroMasters

Week 2 Graded

Assignment

Homework due Mar

02, 2016 at 15:00 UTC

Week 3:

Forecasting II -

MITx: CTL.SC1x Supply Chain Fundamentals

https://d37djvu3ytnwxt.cloudfront.net/asset-v1:MITx+CTL.SC1x_2+1T2016+type@asset+block/Ordroid_Data.odshttps://courses.edx.org/courses/course-v1:MITx+CTL.SC1x_2+1T2016/courseware/aeafb16ac1e64da2a16d98d6cdfa423b/41862172d4ca493996126432b89395ff/https://courses.edx.org/courses/course-v1:MITx+CTL.SC1x_2+1T2016/courseware/aeafb16ac1e64da2a16d98d6cdfa423b/b12ad659313d4f769a575fdc15a43442/https://courses.edx.org/courses/course-v1:MITx+CTL.SC1x_2+1T2016/courseware/aeafb16ac1e64da2a16d98d6cdfa423b/b12ad659313d4f769a575fdc15a43442/https://courses.edx.org/courses/course-v1:MITx+CTL.SC1x_2+1T2016/courseware/aeafb16ac1e64da2a16d98d6cdfa423b/b4a85bbcabc94dc6bf30fe713fb727a1/https://courses.edx.org/courses/course-v1:MITx+CTL.SC1x_2+1T2016/courseware/aeafb16ac1e64da2a16d98d6cdfa423b/fc4eda5f7a9b48a4b77428ea851f328f/https://courses.edx.org/courses/course-v1:MITx+CTL.SC1x_2+1T2016/courseware/aeafb16ac1e64da2a16d98d6cdfa423b/70b1e7abe6cb427fbbea8d16db89ca85/https://courses.edx.org/courses/course-v1:MITx+CTL.SC1x_2+1T2016/courseware/aeafb16ac1e64da2a16d98d6cdfa423b/d4764459b39c40a6b13557e119d73cc2/https://www.edx.org/https://www.edx.org/https://www.edx.org/https://courses.edx.org/courses/course-v1:MITx+CTL.SC1x_2+1T2016/courseware/aeafb16ac1e64da2a16d98d6cdfa423b/41862172d4ca493996126432b89395ff/https://courses.edx.org/courses/course-v1:MITx+CTL.SC1x_2+1T2016/courseware/aeafb16ac1e64da2a16d98d6cdfa423b/b12ad659313d4f769a575fdc15a43442/https://courses.edx.org/courses/course-v1:MITx+CTL.SC1x_2+1T2016/courseware/aeafb16ac1e64da2a16d98d6cdfa423b/b4a85bbcabc94dc6bf30fe713fb727a1/https://courses.edx.org/courses/course-v1:MITx+CTL.SC1x_2+1T2016/courseware/aeafb16ac1e64da2a16d98d6cdfa423b/fc4eda5f7a9b48a4b77428ea851f328f/https://courses.edx.org/courses/course-v1:MITx+CTL.SC1x_2+1T2016/courseware/aeafb16ac1e64da2a16d98d6cdfa423b/70b1e7abe6cb427fbbea8d16db89ca85/https://courses.edx.org/courses/course-v1:MITx+CTL.SC1x_2+1T2016/courseware/aeafb16ac1e64da2a16d98d6cdfa423b/d4764459b39c40a6b13557e119d73cc2/https://d37djvu3ytnwxt.cloudfront.net/asset-v1:MITx+CTL.SC1x_2+1T2016+type@asset+block/Ordroid_Data.odshttps://d37djvu3ytnwxt.cloudfront.net/asset-v1:MITx+CTL.SC1x_2+1T2016+type@asset+block/Ordroid_Data.xlsx


2/7



Answer: 112.5

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


Part 1C

What is the root mean square error (RMSE) of the forecasts in this data

sample?

Answer: 540.6

Exponential

Smoothing

Week 4:

Forecasting III -

Special Cases &

Extensions

Week 5:

Inventory

Management I

- Deterministic

Demand

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.

EXPLANATION

To find this, you just needed to average the absolute values of the

error terms for each observation.


3/7




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


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

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.

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.


4/7



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


Part 2A

Based on your analysis, answer the following questions.

What can you say about the presence seasonality of demand?


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)

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.

https://d37djvu3ytnwxt.cloudfront.net/asset-v1:MITx+CTL.SC1x_2+1T2016+type@asset+block/Ordroid_Solution.odshttps://d37djvu3ytnwxt.cloudfront.net/asset-v1:MITx+CTL.SC1x_2+1T2016+type@asset+block/Ordroid_Solution.xlsx


5/7



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

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

Part 2B

What can you say about the presence of a trend in the demand?


Part 2C

What can you say about the bias of the forecast?

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.


6/7



© All Rights Reserved

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


Part 2D

What can you say about the accuracy of the forecast?


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 notappear to be any bias in the forecast.

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.


7/7



© 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.
https://play.google.com/store/apps/details?id=org.edx.mobilehttps://itunes.apple.com/us/app/edx/id945480667?mt=8http://www.reddit.com/r/edxhttps://plus.google.com/+edXOnlinehttp://www.linkedin.com/company/edxhttps://www.youtube.com/user/edxonline?sub_confirmation=1https://twitter.com/edXOnlinehttp://www.facebook.com/EdxOnlinehttp://open.edx.org/

practice problem 2_ forecasting ordroid devices

Documents