Numerical Algorithms Methods for Computer Vision Machine Learning and Graphics 1st Solomon Solution Manualinstant download
Numerical Algorithms Methods for Computer Vision Machine Learning and Graphics 1st Solomon Solution Manualinstant download
http://testbankbell.com/product/numerical-algorithms-methods-for-
computer-vision-machine-learning-and-graphics-1st-solomon-
solution-manual/
TestBankBell.com: Your Ultimate Source for Test Banks and Academic Resources
Keywords:
test bank, academic resources, study guides, test preparation, testbankbell, download test
bank, textbook solutions
Contact Information:
Visit us at: https://testbankbell.com - For inquiries, email us:
[email protected]
Important Links:
Download Test Banks: https://testbankbell.com/
https://testbankbell.com/product/numerical-methods-and-optimization-
an-introduction-1st-butenko-solution-manual/
testbankbell.com
https://testbankbell.com/product/numerical-methods-in-engineering-
with-matlab-3rd-kiusalaas-solution-manual/
testbankbell.com
https://testbankbell.com/product/solution-manual-for-computer-
graphics-with-open-gl-4-e-4th-edition-0136053580/
testbankbell.com
https://testbankbell.com/product/solution-manual-for-computer-
algorithms-introduction-to-design-and-analysis-3-e-3rd-
edition-0201612445/
testbankbell.com
Solution Manual for Computer Vision: A Modern Approach,
2/E 2nd Edition : 013608592X
https://testbankbell.com/product/solution-manual-for-computer-vision-
a-modern-approach-2-e-2nd-edition-013608592x/
testbankbell.com
https://testbankbell.com/product/solution-manual-for-applied-
numerical-methods-with-matlab-for-engineers-and-science-chapra-3rd-
edition/
testbankbell.com
https://testbankbell.com/product/solution-manual-for-data-structures-
and-algorithms-in-java-1st-edition-peter-drake/
testbankbell.com
https://testbankbell.com/product/solution-manual-for-applied-
numerical-methods-with-matlab-for-engineers-and-scientists-4th-
edition-steven-chapra/
testbankbell.com
https://testbankbell.com/product/solution-manual-for-numerical-
methods-for-engineers-8th-edition-by-steven-chapra-raymond-canale/
testbankbell.com
3
1 2
In×n − A (2In×n − 4A) = In×n − 2A − 2A + 4A
2
= In×n − 2A − 2A + 4A since A is idempotent
= In×n
(d) Aτx = λτx =⇒ A2 τx = λAτx = λ2 τx. But A2 = A, so A2 τx = Aτx = λτx. Thus, λτx = λ2 τx =⇒ λ = λ2 =⇒
λ ∈ {0, 1}.
1.16. The product AB is of size n2 . Obviously finding AB requires considering each element of AB at least once
(if nothing else, to write the result in memory!), already requiring O(n2 ) time even if each element of AB is
computed in O(1) time. The algorithms in the figure take O(n3 ) time to run due to the nested loops. Hence,
there is room for improvement, and indeed Strassen’s algorithm and several others achieve faster than O(n3 )
asymptotic runtime, at least for large n.
1.17. Define λ(τx) ≡ − ln p(τx). Since ln is monotonic, any local maximum of p(τx) is also a local maximum of λ(τx).
Hence, τx∗ is a critical point of λ(τx), implying ∇ λ(τx∗ ) = τ0. Let H be the Hessian of λ at τx∗ . Then, near τx∗ we
can approximate:
1 1
− ln p(τx) = λ(τx) ≈ λ(τx∗ ) + (τx − τx∗ ) H (τx − τx∗ ) = − ln p(τx∗ ) + (τx − τx∗ ) H (τx − τx∗ ).
2 2
The first derivative term of the expansion vanishes since ∇ λ(τx∗ ) = τ0. Exponentiating both sides shows
P̄ V PV V P̄ P
− = −
RT̄ RT R T̄ T
δP T − P δT
=n
P (T + δT )
δP T − P δT T ε P + P εT
≤
P (T + δT ) P (T − εT )
PV (100 Pa)(0.5 m3 )
n= = = 0.0201 mol
RT (8.31 J · mol−1 · K−1 )(300 K)
(101 Pa)(0.5 m3 )
= 0.0203 mol = n + 0.000234 mol = n + 1.17%.
(8.31 J · mol−1 · K−1 )(299.5 K)
The smallest possible value is
(99 Pa)(0.5 m3 )
= 0.0198 mol = n − 0.000234 mol = n − 1.17%.
(8.31 J · mol−1 · K−1 )(300.5 K)
Hence, the absolute error is bounded by 0.0198 mol and the relative error is bounded by 1.17%.
(d) At the range indicated by the problem, it is relatively well-conditioned. When the scale of εT is commen-
surate with that of T , the problem becomes ill-conditioned.
2.3. We can understand the relative error as the fraction
∆y|/|y| x∆y
|
κrel = = ,
|∆x|/|x| y∆x
where y + ∆y = f (x + ∆x) and y = f (x). By Taylor’s theorem, f (x + ∆x) = y + f (x)∆x + O(∆x2 ). Hence,
∆y = f (x)∆x + O(∆x2 ), so for small ∆x,
x · f (x)∆x xf (x)
κrel ≈ = .
f (x) · ∆x f (x)
∆y
≈ |f (x)| .
∆x
The function f (x) = ln x has a large relative condition number near x = 1, since κrel = 1/ln x, which blows up
near x = 1. Contrastingly, the function f(x) = x has relative condition number 1 for all x.
2.4. Since minima are roots of f , we can use the conditioning for root-finding, but with an extra derivative:
(a) |xest − x∗ |
(b) |f (xest ) − f (x∗ )| ≈ δx|f (x∗ )|
2.5. (a) The range is (−∞, 0] since limt→0 log t = −∞ and log 1 = 0.
(b) If the xk is very negative, then exk is exponentially close to zero. This near-zero value may not be repre-
sentable, and regardless a single slightly larger value will dominate the sum.
χ(x1 , . . . , xn ) = ln e xk by definition
k
= ln exk −a+a
k
= ln ea exk −a
k
= ln ea + ln exk −a
k
= a + ln exk −a
k
Suppose we take a = mink xk . Then, rather than adding together tiny values we have moved the scale to
be around e0 = 1. (Other heuristics for choosing a are possible)
2.6. There are rendering artifacts because the two surfaces overlap and hence have the same depth values; rounding
during depth computation can make one surface appear on top of the other. Possible resolutions include
slightly offsetting one surface, adding a tie-breaking rule when depths are within some tolerance of each other,
or merging the geometry before rendering to avoid overlap altogether.
2.7. (a) Recall that floating point arithmetic changes spacing as the order of magnitude of the value changes. Thus,
it makes sense to have multiplicative error that is relative to the scale of x and y.
(b) (adapted from course notes by D. Bindel, Cornell CS) The recurrence for the ground-truth sum is simply
sk = sk−1 + xk yk . Error terms for the addition and multiplication steps show
ŝk = (ŝk−1 + xk yk (1 + ε×k ))(1 + εk+).
Subtracting the two shows:
= x1 y1 (ε+ × + + × + + + × + × + ×
1 + ε1 + ε1 ε2 + ε1 ε2 + ε 2 ) + x2 y2 (ε 2 + ε 2 ) + [x1 y1 ε 1 ε 1 + x2 y 2 ε 2 ε 2 ] + O(ε max)
3
ŝk − sk = xi yi εi× + ε+
j
2
+ O(kεmax)
i=1 j=i
2.8. For convenience, define d ≡ x− y. We’ll start by simplifying the numerator of relative error and then substitute:
(1 + εx )x − (1 + εy )y = (x − y) + (εx x − εy y)
= d + εx d + (εx − εy )y
=⇒ (1 + ε− )((1 + εx )x − (1 + εy )y) = (1 + ε− )(d + εx d + (εx − εy )y)
= (1 + ε− )d + εx (1 + ε− )d + (1 + ε− )(εx − εy )y
(1 + ε− )((1 + εx )x − (1 + εy )y) − (x − y)
=⇒ E =
x− y
ε− d + εx (1 + ε− )d + (1 + ε− )(εx − εy )y
=
d
y
=
ε− + εx (1 + ε− ) + (1 + ε− )(εx − εy )
d
This can be unbounded as d → 0.
2.9. (a) Implicitly differentiating the relationship 0 = f (x(ε)) + εp(x(ε)) with respect to ε shows
d
0= [f (x(ε)) + εp(x(ε))]
dε
= f (x(ε))x (ε) + p(x(ε)) + εp (x(ε))x (ε) by the chain rule.
p(x∗ )
0 = f (x∗ )x (0) + p(x∗ ) =⇒ x (0) = − .
f (x∗ )
(b) We differentiate
d
f (x) = (x − 1) · (x − 2) · · · · · (x − 20)
dx
= (x − 2) · · · · · (x − 20) + (x − 1) · (x − 3) · · · · · (x − 20)
+ · · · + (x − 1) · · · · · (x − 19) by the product rule
Substituting x = j shows
(c) x (1) ≈ 8.2 × 10−18 and x (20) ≈ −4.3 × 107 ; hence, the root x∗ = 1 is far more stable.
2.10. (a) The alternative formula can be obtained by scaling the numerator and denominator of the quadratic
equation:
√ √ √
−b ± b2 − 4ac −b ± b2 − 4ac −b ∓ b2 − 4ac
= · √
2a 2a −b ∓ b2 − 4ac
b2 − (b2 − 4ac)
= √
−2ab ∓ 2a b2 − 4ac
4ac
= √
−2ab ∓ 2a b2 − 4ac
−2c
=
This way, there never can be cancellation because we always move b farther from the origin in the numer-
ator.
2.11. The bounds are worked out below:
[x] + [y] = [x + y, x + y]
[x] − [y] = [x − y, x − y]
value sign(x) sign(x) sign(y) sign(y)
xy, xy + + + +
xy, xy + + − +
yx, yx + + − −
[xy, xy] − + + +
[x] × [y] =
min(xy, yx), max(xy, xy) − + − +
xy, x − + − −
xy, xy − − + +
[xy, xy] − − − +
xy, xy − − − −
1 1
[x] ÷ [y] = [x] × ,
y y
[x]1/2 = [x1/2 , x1/2 ]
In finite-precision arithmetic, always round down the lower bounds and round up the upper bounds.
2.12. (a) Perturbing any of three collinear points slightly makes them not collinear. Furthermore, points may appear
collinear if you zoom out far enough but appear less so as you zoom in.
ε
q
ε
p
ε
ε
(b)
ε
φq
ε
pφ
ε
(c)
(d) Obvious from drawings above; ε-collinear points form the intersection of four half-planes, two of which
come from the ε-clockwise condition and two of which come from the ε-counterclockwise condition.
(e) No. See §3.1 of [55] for an example.
3.1. No; LU may not be possible for matrices requiring pivoting.
3.2. The steps of Gaussian elimination are below:
2 4 2 1 2 1 1/2 0
∼ , with elimina tion matrix
3 5 4 3 5 4 0 1
1 2 1 1 0
∼ , with elimination matrix
0 1 −1 3 −1
1 0 3 1 −2
∼ , with elimination matrix
0 1 −1 0 1
−1 −1
1/2 0 1 0 2 0 1 0 2 0
L= = = .
0 1 3 −1 0 1 3 −1 3 −1
1 0 0 1 2 7
L= 3 1 0 U= 0 −1 −22
6 11 1 0 0 204
3.4. Where it states “optionally insert pivoting code here,” find row r with largest value in column p; then swap
row r and row p of both A and φb.
3.5. No. Full pivoting can be preferable numerically but technically does not make a difference. The only way partial
pivoting would fail is if there is an all-zero column, which would indicate that A is not invertible.
3.6. Write A = A1 + A2 i, φb = φb1 + φb2 i, and φx = φx1 + φx2 i. Then, Aφx = φb =⇒ (A1 + A2 i)(φx1 + φx2 i) = φb1 + φb2 i =⇒
(A1 φx1 − A2 φx2 ) + (A2 φx1 + A1 φx2 )i = φb1 + φb2 i. So, we can solve the block system
A2
A1 −A2 A1 φx1
φb1 = .
φ
b
2
3.7. Carrying out Gaussian elimination is the same as pre-multiplying by the inverse of the leftmost n × n block.
Hence, the output is A−1 (A|In ×n ) = (A−1 A|A−1 ) = (I n×n |A−1 ).
“1. For taking God’s blessing out of the Sea to enrich the
Realm, which otherwise we lose.
2. For setting the People on work.
3. For making Plenty of Cheapness in the Realm.
4. For increasing of Shipping, to make the Land powerful.
5. For a continual Nursery for breeding and increasing our
Mariners.
6. For making employment of all Sorts of People, as blind,
lame, and others, by Sea and Land, from ten or twelve years
and upwards.
7. For inriching your Majesty’s Coffers, by Merchandises
returned from other Countries for Fish and Herrings.
8. For the increase and enabling of Merchants, which now
droop and daily decay.”
The completion of the New River in 1620 was a great boon and
blessing to the people, but the greatest benefit to trade in the reign
of James I. was the improvement of the navigation of the upper part
of the Thames by deepening the channel, so that not only was
Oxford placed in communication with London, but the country all
round Oxford.
The granting of monopolies was an interference with trade which
would now cause a revolution. There were many complaints.
Parliament declared that all monopolies were void. That was under
James. Charles began, notwithstanding, to sell monopolies to
whomsoever would pay him most for them. Thus the importation of
alum was prohibited, for the protection of the alum works of Whitby;
also brick-making, the manufacture of saltpetre, of tapestry, the
coining of farthings, the making of steel, the making of stone pots
and jugs, making guns, melting iron ore, and many other things.
More than this, Charles made the sale of tobacco a royal monopoly;
he forbade the infant colony of Virginia to sell tobacco to any foreign
state; he levied a duty of four shillings a chaldron on all coal
exported to foreign parts; and he actually endeavoured to establish
a malting and a brewing monopoly. When we read the historian on
the despotic acts of Charles and his attempts on the liberties of his
people, let us bear in mind the constant exasperations of these
interferences with trade—that is, with the livelihood of the people.
When at last he became awakened to the danger of the position, he
revoked all their “grants, licences, and privileges”; but it was then
too late—revolution had already arisen.
Shops which had been open stalls confined to one or two markets
in London, such as East and West Chepe, began, towards the end of
the sixteenth century and early in the seventeenth, to appear along
Fleet Street, the Strand, and in King Street, Westminster.
Haberdashers, milliners, woollen drapers, cutlers, upholsterers,
glassmen, perfumers, and others established themselves
everywhere, making so brave a show every day, that, as Stow
complains, “the people of London began to expend extravagantly.”
There were offered, among other wares, “French and Spanish
gloves, and French cloth or frigarde (frieze), Flanders-dyed kersies,
daggers, swords, knives, Spanish girdles, painted cruses, dials,
tables, cards, balls, glasses, fine earthen pots, salt-cellars, spoons,
tin dishes, puppets, pennons, ink-horns, toothpicks, silk, and silver
buttons. All which ‘made such a show in passengers’ eyes, that they
could not help gazing on and buying these knicknacks.’ This great
offence a contemporary writer, quoted by Stow, bitterly
apostrophises. He ‘marvels’ that ‘no man taketh heed to it what
number of trifles cometh hither from beyond the seas, that we might
either clean spare, or else make them within our own realm; for the
which we either pay inestimable treasure every year, or else
exchange substantial wares and necessaries for them, for the which
we might receive great treasure.”
There had then arisen outside the City a new class, and one which
was becoming wealthy and important, namely, the suburban
shopkeepers. They were certainly not a class that Charles could
afford to exasperate. But apparently he never asked himself how far
it was prudent to exasperate any class. Thus, in the blindness of his
wrath against the Puritans, whose emigration was the best thing
that could happen to him, he forbade them to emigrate without a
certificate of having taken the oath of allegiance and supremacy, and
likewise from the minister of their parish a certificate of their
conversation and conformity to the orders and discipline of the
Church of England. He therefore did what he could to preserve his
own enemies in his kingdom, and to increase their hostility. Again,
he ordered that the Weavers’ Company should admit to its freedom
none but members of the Church of England. He even interfered
with trade to the extent of trading on his own account, on one
occasion buying up all the pepper imported by the East India
Company and selling it again at a profit.
The foundation of the banking business is said to date from the
outbreak of the Civil War; perhaps it was partly due to that event.
Banking was impossible in earlier times for several reasons: there
was no system of commercial credit; there were no bank-notes;
goods were bought or sold for actual coin; there was no Exchange;
when men went abroad or came home, they had to take their
foreign money to the Mint for re-coinage, or they had to get foreign
money at the Mint; there was no recognised system of lending or
borrowing; if a man borrowed money he did so as a special occasion
and for a special purpose, and paid a large interest for the
accommodation. The money-lenders were the Jews first, who carried
on the trade as a Royal monopoly, followed by the Lombards, who
came as the agents of Papal taxation; and afterwards the London
merchants and goldsmiths.
When the Civil War broke out it became a serious consideration
with the merchants to place their money in some place of security.
The Mint, their former place of deposit, could not be trusted because
Charles had already seized upon £200,000 belonging to merchants,
and placed there for safety; their own strong rooms would not do,
because if the City fell into the hands of the Royalists, the strong
room would most certainly be plundered first.[8] They therefore
began to lodge their cash in the hands of goldsmiths, keeping what
was called a “running cash” account. They probably thought that in
case of need the goldsmiths could take their money and plate
abroad. Country gentlemen also began to send their money up to
London for greater security. This method was found so convenient
that banking quickly spread and the bankers began to flourish.
During the Commonwealth one Henry Robinson proposed the
establishment of a “Land Bank,” with branches in the country to lend
money upon mortgage, the payments to be by paper.
THE FIRST ROYAL EXCHANGE—EXTERIOR AND INTERIOR
The most important civic event in the reign of William was the
foundation of the Bank of England. The bankers of the City had
been, as we have seen, the goldsmiths, who not only received and
kept money for their customers, but lent it out, for them and for
themselves, on interest. Thus, when Charles the Second, on January
2, 1672, closed the Exchequer, his creditors could obtain neither
principal nor interest. He then owed the goldsmiths of London the
sum of £1,328,526. Some of the older banks of London trace a
descent to the goldsmiths of the seventeenth century—Child’s, for
instance, Hoare’s, and others.
A PERSPECTIVE VIEW OF THE BANK OF ENGLAND, 1743
From Maitland’s History and Survey of London.
The last loan ever asked by the King of the City was that asked by
William the Third in 1697 to pay off his navy after the Peace of
Utrecht. Instead of going to the City the King went to the Bank, but
not then without the authority of Parliament. The City was no longer
to be the Treasurer—or the Pawnbroker—of the King and the nation.
This event, though the citizens did not apprehend its full meaning
for many years, deprived London of that special power which had
made her from the Norman Conquest alternately the object of the
Sovereign’s affections or of his hatred. Henceforth it mattered
nothing to London whether the King loved or hated her. The power
of the City was now exercised legitimately by her representatives in
the House of Commons.
CH APTER VII
THE IRISH ESTATES
The following history of the Irish estates is taken from London and
Londonderry, published in Belfast (1890) by Messrs. Marcus Ward
and Co.
In the year 1608 the first steps were taken towards the settlement
of Ulster by English and Scotch emigrants—a measure whose
wisdom was shown eighty years later, when the grandson of James
the First owed his expulsion largely to the descendants of the
original settlers.
The greater part of six counties in the Province of Ulster, viz.
Donegal, Fermanagh, Cavan, Tyrone, Armagh, and Coleraine, after
the rebellions of O’Neill and O’Donnell, were declared to be
escheated to the Crown. James conceived a plan for securing the
peace and welfare of Ulster by replacing the Irish rebels by
Protestant settlers, together with those of the Irish who were willing
to conform to the English rule and religion. He therefore invited
“undertakers” who would accept of lands in Ulster on his conditions.
These were, that they should not ask for large portions “in tending
their private property only;” that there should be three classes of
undertakers:
(1) Those who would plant with English or Scotch tenants.
(2) Servitors or military undertakers.
(3) Native Irish admitted as freeholders.
The first class were to pay to the Crown the great rent of £5:6:8
for every thousand acres. The second class, when they planted with
English or Scotch tenants, were to pay the same; otherwise, the
second class were to pay £8 for every thousand acres; and the third
class were to pay £10:13:4 for every thousand acres.
The third condition was, that all were required to provide
strongholds and arms for defence, to let their lands on easy terms,
to “avoid Irish exactions,” and to be resident; they were not to
accept the “mere” Irish as tenants at all; they were required to
create market towns, and to found at least one free school in every
county for education in religion and learning. They were also
privileged to import from Great Britain for three years, free of
custom, everything requisite to put the plantation on a satisfactory
footing.
In 1609 Commissioners were appointed to survey the escheated
lands and to divide them into convenient parcels for allocation.
In the same year proposals were made in the King’s name to the
City that the Corporation itself should undertake the restoration of
the city of Derry and the town of Coleraine, and should plant the
rest of the county with undertakers. The City was offered the
Customs for twenty-one years at 6s. 8d. per annum, the fisheries of
the Bann and the Foyle, free license to export wares grown on their
own land, and the admiralty of Tyrconnel and Coleraine.
The following were the inducements held out to the City:—
On March 29, 1613, the first charter was granted to the Irish
Society as representing not the Companies, but the Corporation of
the City of London for the Plantations. This charter constitutes and
incorporates the Irish Society:—
“‘For the better ordering, directing, and governing all and all
manner of things for and concerning the City and Citizens of
Londonderry aforesaid, and the aforesaid County of
Londonderry, and the Plantation to be made within the same
City and County of Londonderry, and other businesses belonging
to the same,’ giving the Society power to purchase and hold in
fee, for these purposes, lands, goods, etc., in England or in
Ireland, to have a common seal, and to sue or be sued” (p. 14).
A grant of timber is made only for the Plantation and “not for any
other causes to be merchandized or sold.” In other words, the Irish
Society was incorporated for the purpose of a Trust; the members
were originally Trustees.
It is charged against the Society that they began by neglecting the
conditions, setting too high a rent upon their lands, and trying to
make a profit for the Londoners out of the property. James himself
was much dissatisfied with the conduct of the estates. He wrote to
Sir Arthur again in August 1615, adding a postscript in his own hand:
—“My Lord, in this service I expect that zeal and uprightness from
you, that you will spare no flesh, English or Scotch, for no private
man’s worth is able to counterbalance the particular safety of a
kingdom, which this Plantation, being well accomplished, will
procure.”
A year afterwards the King granted a licence to the twelve
Companies to hold in mortmain whatever lands the Irish Society
might grant them.
These grants contained a reservation of the right of re-entry if the
conditions specified were not kept.
It is, of course, evident that if the Irish Society were Trustees they
could not give away their lands, and that they could only make
grants under the conditions of their Trust.