OUTLINES OF SOLUTIONS
‘The following are outlines of solutions extracted from the annual reports of past Hong
examinations. Readers should note that they are not meant to
outline of Solutions
1080
Paper 
w
Using properties of matrices, we have AO = 0, and for any
x. ¥ &E, andany a, BER,
Alax + By) = @(Ax)+ play) = 0
Hence 0 € E, andforany x, y €E, andany a, BER
ax + By © E, showingthat &
Wey = ptx with xEE, then Ay = Alp + x) =
Ag+ Ax=b . Ontheatherhand,if Ay = b, then
putting x= y~p, wehae Ax= AyAp=0,
showing xE EF and y=ptx.
By routine method,
solutionset = [(3#, 2e, 6r) : 2 © R}
Applying (aNii) and (b)(i, A
solution set = [( 4 3¢, 420, V2 +60
Reflexive: Let I(x) 21.
Forany FEF, 1+ f= £* 1 and I isa polynomial,
therefore f~ 7
Symmetric: If f ~ g, then there exist polynomials p, q © F such
Transitive: tf ~ g and g ~ h, then th
that
pef=aqegorg
ont.
pet
P. a. 1, & © F suchthat
psf a9, rege seh
Je and associative, we get
+8) eh whee rep and ges
in F. Thus f~
Noting + is commuta
(replete
are both polynomiWty ~ = fi/~ and g/~ = g/~, then 03 (ch Let x, = 2% then x, > 0 and Bx, =n.
Deft aeh nd rege sey
for some polynomials p, @, r, § € F. Then
(perle(feg) = (aesde th oy) +rm
So > (6 oo + nf
if the equality holds and m #1, then
where p +f and q+ § are polynomials in F. Thus
Fag ~ foo, ie fe gl~ = Uh eal/~ oF
t/~@ 9!~ = fhf~ OH/~
Tq Ny giving Y= ey,
(i) Closure: follows from (by).
Associatvity: follows from the definition of @ and the
associativity of + be ,
lentiy: ——1/~ nteidentiy since torany VE F/~, 0k ty AR et at
ti~@I/~ = (fe Ii~ = f/~, and
Tae ee Te This formula can be established by mathematical induction,
Inverse: For any f/~ © F/~, let #7) be the inverse (0) x, — an = oy a
of f in F. Then Gb Load *
f[~O I'm = 1/~, = 2, Ahan y 1 = 80 Where 2 XE OK yy
fs @t/~ = Tm Bok ome oe
fependent of k and 2% 1. From (a),
Poy =e
Fu, a= omy 7G
Differentiating fx) = S2*L= 1  xP — 1 we find that the
pti Pp
absolute minimum of f on (0, =) is O et x = 1. Hence for all x > 0,
xPthaa
ptt
and the equality holds only if x = 1.
Take
By induction om and using (a).
W xy, # 1 for some fp , then by (i Q5 (2) (i) By substitution, we find that u+v, wu tw, wutuv
5 ‘are roots of the equation
2 duvx = +P) 0
i Consider
+e =6 e+e =6
un? 8 lee as 1
Werfind that u? = 2 and v? = 4. Asolutionis v= 2? ,
imptving es
ea ene “
1 1
wa > ies >o
™ mat ae
oe Bast >n, acontadicion
Aas
as
a7
(a)
to)
Cc)
@
o
w
i
Note
Expand the rightha
f

This
o
Let a, @, B be the roots of (+). By the relations between roots
and coetficients, we get
Qatp=0, a +2ab=p, a Bra
Eliminating @ and B, weget 2797 + 4p? =
The rootsof (+) ae uty, wut uty, wry + wy, where
w+Pea.
1
wrote
whien has u =v =~ JZ svasotutionit 2707 + 49° = 0°
Thus wu + wy = wu + wy isamultiple oot of (+)
mth etx ~ alu +(x  ym”
that [It + ux (au + 6) \
side and compare the coefficients of u’
mini(a~bym* nth
(m+a+ii
“=a (x = by = (=
a.
(e_ym enn
it mmtn or m>r.
can be easily verified
From the properties of g we get
io(zl = iz, Ite) t = lent
Using properties of complex numbers, we have
gizil? = (g(z)+ acai) +1
= edad
Therefore b= £1.
a7
as
@
(al
3)
=i, then
Thus we get
als =i,
ioiz) +
Weget 912)  giz) = 27. By
,olzyaz.
= fiz) = £10) Saat
9, given by g (2) = FEL —ETEY isan isometry satisfying th
conditions in (b), Then f(z) = ag(z) +b with
a= f(1)~ (0) and b= FLO)
Thus Lal = 1. By (b) and (c), wehave
fz) = ert o flz)= Fob
te”. Ne
GM in = 1
y ok
Since ko is most probable, we have
praetor)
By computation we get
Nn+1
wey
aS KOS
Mean number =
k
Pa i ae 2)
Differentiating both sides of
io
(p + qx) = Fiqve AK gait
with eect to x antenna x 23, 9 =
qe We we get
mean nber =
since (11) chy Mot stmeretoreto ~ AL 1 .
‘Take out one cell and consider this to be occupied by exactly k balls
(1S SN). There are G.% ways of doing so, Then the number of
ways such that the remining cells being occupied by WV ~k_ balls is
AIN =k, 0). Thus
W
Ain, nen = 2 Gan —k,
The second partis proved by induction on. For = 1, itcan be
emily verified, Assume i