Professional Documents
Culture Documents
Solution Manual For Forecasting and Predictive Analytics With Forecast X 7th Edition by Keating
Solution Manual For Forecasting and Predictive Analytics With Forecast X 7th Edition by Keating
Solution Manual For Forecasting and Predictive Analytics With Forecast X 7th Edition by Keating
For a supply chain to function efficiently all the participants along the chain must have accurate
forecasts in order to plan production and distribution. Consider a firm that makes seats for a
major auto manufacturer. This firm needs a reliable forecast of the production of vehicles so that
they can plan their raw material and human resource requirements.
5. The process of forecasting new products is difficult. Why? How can new products be
forecast?
The biggest hurtle to overcome in new product forecasting is the lack of historical data. In the
most extreme case of a totally new product there is no historical data that can be used to help
develop a forecast. Often a new product is not completely new to the market. There may be
similar products on the market for which there is a body of data that might be useful. Consider
the newest iPhone. Apple has historical data on previous versions of the iPhone that can be a
guide to expected sales of the newest model. When a product is totally new a firm can rely on
marketing research in the form of test markets and product clinics to develop estimates of future
sales. Once the product is introduced and a few data points have been observed the Bass model
may be useful. Products typically follow a product life cycle which provides a framework for
developing sales estimates.
6. In this chapter, you saw an example of a naive forecast. Why do you think it is given that
name? Describe how the naive forecast is developed.
Naïve forecasts are given that name because they simply assume that the next period will be the
same as the current period. This is truly a naïve assumption. Most people use a naïve forecast in
their daily lives. In the absence of a weather forecast, it is not uncommon to think that the
weather tomorrow will be the same as it is today. Most of the time this works well enough in the
very short term but such a forecast becomes absurd in a longer term. It is likely that the weather
on January 15 will be similar to the weather January 14th. But extending that reasoning six
months ahead the June 15 would be unreasonable. In the simplest form a naïve forecast takes
today’s sales as the forecast for tomorrow’s sales. That is the forecast for time period t+1 is the
actual value at time period t.
MAPE stands for “Mean Absolute Percentage Error.” The MAPE is one form of an average
error in a forecast over a specific period of time. Because the absolute value of the errors are
used in the calculation it is not possible for positive errors to offset negative errors. Suppose you
have a forecast for four quarters with errors of 100. 50, -225, and 175. Calculating a simple
average would tell you that on average there was zero error. In fact this is quite obviously not
true. For simplicity assume that actual sales for the four periods were as shown in the table
below. The MAPE would be calculated as shown below:
Absolute Absolute
Actual Forecast Error Error % Error
400 300 100 100 25.00
200 150 50 50 25.00
500 725 -225 225 45.00
400 225 175 175 43.75
MAPE = 34.69
The MAPE for the same four periods would be 34.69 which would be a better indication of the
true errors in the forecast and would not mislead a manager into thinking the forecasts had all
been perfect. MAPE is also unit free since it is calculated as a percent. Thus, the MAPE can be
compared across different forecast items.
8. Suppose that you work for a U.S. senator who is contemplating writing a bill that would
put a national sales tax in place. Because the tax would be levied on the sales revenue of
retail stores, the senator has asked you to prepare a forecast of retail store sales for year 8,
based on data from year 1 through year 7. The data are:
1. Use the naive forecasting model presented in this chapter to prepare a forecast of retail
store sales for each year from 2 through 8.
2. Prepare a time-series graph of the actual and forecast values of retail store sales for the
entire period. (You will not have a forecast for year 1 or an actual value for year 8.)
3. Calculate the MAPE for your forecast series using the values for year 2 through year 7.
The troubles in the South, and the almost general overthrow of the
“carpet-bag government,” impressed all with the fact that the
Presidential election of 1876 would be exceedingly close and exciting,
and the result confirmed this belief. The Greenbackers were the first
to meet in National Convention, at Indianapolis, May 17th. Peter
Cooper of New York was nominated for President, and Samuel F.
Cary of Ohio, for Vice-President.
The Republican National Convention met at Cincinnati, June 14th,
with James G. Blaine recognized as the leading candidate. Grant had
been named for a third term, and there was a belief that his name
would be presented. Such was the feeling on this question that the
House of Congress and a Republican State Convention in
Pennsylvania, had passed resolutions declaring that a third term for
President would be a violation of the “unwritten law” handed down
through the examples of Washington, and Jackson. His name,
however, was not then presented. The “unit rule” at this Convention
was for the first time resisted, and by the friends of Blaine, with a
view to release from instructions of State Conventions some of his
friends. New York had instructed for Conkling, and Pennsylvania for
Hartranft. In both of these states some delegates had been chosen by
their respective Congressional districts, in advance of any State
action, and these elections were as a rule confirmed by the State
bodies. Where they were not, there were contests, and the right of
district representation was jeopardized if not destroyed by the
reinforcement of the unit rule. It was therefore thought to be a
question of much importance by the warring interests. Hon. Edw.
McPherson was the temporary Chairman of the Convention, and he
took the earliest opportunity presented to decide against the binding
force of the unit rule, and to assert the liberty of each delegate to vote
as he pleased. The Convention sustained the decision on an appeal.
Ballots of the Cincinnati Republican Convention, 1876:
Ballots, 1 2 3 4 5 6 7
Blaine, 285 296 292 293 287 308 351
Conkling, 113 114 121 126 114 111 21
Bristow, 99 93 90 84 82 81
Morton, 124 120 113 108 95 85
Hayes, 61 64 67 68 102 113 384
Hartranft, 58 63 68 71 69 50
Jewell, 11
Washb’ne, 1 1 3 3 4
Wheeler, 3 3 2 2 2 2
The election followed Nov. 7th, 1876, Hayes and Wheeler carrying
all of the Northern States except Connecticut, New York, New Jersey
and Indiana; Tilden and Hendricks carried all of the Southern States
except South Carolina, Florida and Louisiana. The three last named
States were claimed by the Democrats, but their members of the
Congressional Investigating Committee quieted rival claims as to
South Carolina by agreeing that it had fairly chosen the Republican
electors. So close was the result that success or failure hinged upon
the returns of Florida and Louisiana, and for days and weeks
conflicting stories and claims came from these States. The Democrats
claimed that they had won on the face of the returns from Louisiana,
and that there was no authority to go behind these. The Republicans
publicly alleged frauds in nearly all of the Southern States; that the
colored vote had been violently suppressed in the Gulf States, but
they did not formally dispute the face of the returns in any State save
where the returning boards gave them the victory. This doubtful
state of affairs induced a number of prominent politicians of both the
great parties to visit the State capitals of South Carolina, Florida and
Louisiana to witness the count. Some of these were appointed by
President Grant; others by the Democratic National Committee, and
both sets were at the time called the “visiting statesmen,” a phrase on
which the political changes were rung for months and years
thereafter.
The electoral votes of Florida were decided by the returning board
to be Republican by a majority of 926,—this after throwing out the
votes of several districts where fraudulent returns were alleged to be
apparent or shown by testimony. The Board was cited before the
State Supreme Court, which ordered a count of the face of the
returns; a second meeting only led to a second Republican return,
and the Republican electors were then declared to have been chosen
by a majority of 206, though before this was done, the Electoral
College of the State had met and cast their four votes for Hayes and
Wheeler. Both parties agreed very closely in their counts, except as to
Baker county, from which the Republicans claimed 41 majority, the
Democrats 95 majority—the returning board accepting the
Republican claim.
In Louisiana the Packard returning board was headed by J.
Madison Wells, and this body refused to permit the Democrats to be
represented therein. It was in session three weeks, the excitement all
the time being at fever heat, and finally made the following average
returns: Republican electors, 74,436; Democratic, 70,505;
Republican majority, 3,931. McEnery, who claimed to be Governor,
gave the Democratic electors a certificate based on an average vote of
83,635 against 75,759, a Democratic majority of 7,876.
In Oregon, the three Republican electors had an admitted majority
of the popular vote, but on a claim that one of the number was a
Federal office-holder and therefore ineligible, the Democratic
Governor gave a certificate to two of the Republican electors, and a
Mr. Cronin, Democrat. The three Republican electors were certified
by the Secretary of State, who was the canvassing officer by law. This
Oregon business led to grave suspicions against Mr. Tilden, who was
thereafter freely charged by the Republicans with the use of his
immense private fortune to control the result, and thereafter, the
New York Tribune, with unexampled enterprise, exposed and
reprinted the “cipher dispatches” from Gramercy, which Mr. Pelton,
the nephew and private secretary of Mr. Tilden, had sent to
Democratic “visiting statesmen” in the four disputed sections. In
1878, the Potter Investigating Committee subsequently confirmed
the “cipher dispatches” but Mr. Tilden denied any knowledge of
them.
The second session of the 44th Congress met on Dec. 5th, 1876,
and while by that time all knew the dangers of the approaching
electoral count, yet neither House would consent to the revision of
the joint rule regulating the count. The Republicans claimed that the
President of the Senate had the sole authority to open and announce
the returns in the presence of the two Houses; the Democrats plainly
disputed this right, and claimed that the joint body could control the
count under the law. Some Democrats went so far as to say that the
House (which was Democratic, with Samuel J. Randall in the
Speaker’s chair) could for itself decide when the emergency had
arrived in which it was to elect a President.
There was grave danger, and it was asserted that the Democrats,
fearing the President of the Senate would exercise the power of
declaring the result, were preparing first to forcibly and at least with
secrecy swear in and inaugurate Tilden. Mr. Watterson, member of
the House from Kentucky, boasted that he had completed
arrangements to have 100,000 men at Washington on inauguration
day, to see that Tilden was installed. President Grant and Secretary
of War Cameron, thought the condition of affairs critical, and both
made active though secret preparations to secure the safe if not the
peaceful inauguration of Hayes. Grant, in one of his sententious
utterances, said he “would have peace if he had to fight for it.” To this
end he sent for Gov. Hartranft of Pennsylvania, to know if he could
stop any attempted movement of New York troops to Washington, as
he had information that the purpose was to forcibly install Tilden.
Gov. Hartranft replied that he could do it with the National Guard
and the Grand Army of the Republic. He was told to return to
Harrisburg and prepare for such an emergency. This he did, and as
the Legislature was then in session, a Republican caucus was called,
and it resolved, without knowing exactly why, to sustain any action of
the Governor with the resources of the State. Secretary Cameron also
sent for Gen’l Sherman, and for a time went on with comprehensive
preparations, which if there had been need for completion, would
certainly have put a speedy check upon the madness of any mob.
There is a most interesting unwritten history of events then
transpiring which no one now living can fully relate without
unjustifiable violations of political and personal confidences. But the
danger was avoided by the patriotism of prominent members of
Congress representing both of the great political parties. These
gentlemen held several important and private conferences, and
substantially agreed upon a result several days before the exciting
struggle which followed the introduction of the Electoral
Commission Act. The leaders on the part of the Republicans in these
conferences were Conkling, Edmunds, Frelinghuysen; on the part of
the Democrats Bayard, Gordon, Randall and Hewitt, the latter a
member of the House and Chairman of the National Democratic
Committee.
The Electoral Commission Act, the basis of agreement, was
supported by Conkling in a speech of great power, and of all men
engaged in this great work he was at the time most suspected by the
Republicans, who feared that his admitted dislike to Hayes would
cause him to favor a bill which would secure the return of Tilden, and
as both of the gentlemen were New Yorkers, there was for several
days grave fears of a combination between the two. The result
showed the injustice done, and convinced theretofore doubting
Republicans that Conkling, even as a partisan, was faithful and far-
seeing. The Electoral Commission measure was a Democratic one, if
we are to judge from the character of the votes cast for and against it.
In the Senate the vote stood 47 for to 17 against. There were 21
Republicans for it and 16 against, while there were also 26
Democrats for it to only 1 (Eaton) against. In the House much the
same proportion was maintained, the bill passing that body by 191 to
86. The following is the text of the
“Geo. P. Rarey.”
Wilkinson Call.”
Also the following:
“Tallahassee, December 4, 1876.
“W. T. Pelton:
W. Call.”
OREGON.
“Gov. L. F. Grover:
“W. H. Effinger,
“A. Noltner,
“C. P. Bellinger.”
“We want to see you particularly on account of despatches from the East.
“William Strong,
“C. P. Bellinger,
“S. H. Reed,
“W. W. Thayer,
“C. E. Bronaugh.”
Also the following cipher despatch sent from Portland, Oregon, on
the 28th day of November, 1876, to New York City:
“Portland, November 28, 1876.
J. H. N. Patrick.
“James K. Kelly.”
Of which, when the key was discovered, the following was found to
be the true intent and meaning:
“Portland, November 28, 1876.
J. H. N. Patrick.
“James K. Kelly.”
C. E. Tilton.”
“I shall decide every point in the case of post-office elector in favor of the highest
Democratic elector, and grant certificate accordingly on morning of 6th instant.
Confidential.
Governor.”
“Funds from New York will be deposited to your credit here to-morrow when
bank opens. I know it. Act accordingly. Answer.
W. C. Griswold.”
[No signature.]
“No. How soon will Governor decide certificate? If you make obligation
contingent on the result in March, it can be done, and slightly if necessary.”
[No signature.]
“Governor all right without reward. Will issue certificate Tuesday. This is a
secret. Republicans threaten if certificate issued to ignore Democratic claims and
fill vacancy, and thus defeat action of Governor. One elector must be paid to
recognize Democrat to secure majority. Have employed three lawyers, editor of
only Republican paper as one lawyer, fee $3,000. Will take $5,000 for Republican
elector; must raise money; can’t make fee contingent. Sail Saturday. Kelly and
Bellinger will act. Communicate with them. Must act promptly.”
[No signature].
“J. H. N. Patrick.”
Kountze Bros.”
[No signature.]
“Is your matter certain? There must be no mistake. All depends on you. Place no
reliance on any favorable report from three southward. Sonetter. Answer quick.”
[No signature.]
“Glory to God! Hold on to the one vote in Oregon! I have one hundred thousand
men to back it up!
“Corse.”
Wm. P. Frye.
O. D. Conger.
E. G. Lapham.
Leave was given to Mr. Knott to present his individual views, also
to Mr. Butler (the full committee consisting of Messrs. Knott,
Lynde, Harris, of Virginia, Hartridge, Stenger, McMahon,
Culberson, Frye, Butler, Conger, Lapham.)
The question being on the resolution reported by the committee, it
was agreed to—yeas 235, nays 14, not voting 42.
The Hayes Administration.