2024 HSC Mathematics Extension 2

Tackling Proofs
‼️ This is a document with an authorial bias as it contains notes regarding his own thinking process.

amiyuki7
Last Resort Clutch Up Algebraic Inequalities
Most Extension 2 algebraic inequalities (which don't necessarily require you to consider the sum or
product of other inequalities) can be solved by the following useful inequalities:
1. HM⩽GM⩽AM⩽QM (harmonic, geometric, arithmetic, quadratic mean chain)
This "mean inequality chain" is deduced by proving the AM/GM inequality, then proving the HM/GM
and AM/QM inequalities, then combining by the transitive property of inequalities (a b c a c . < < ⟹ < )

For all xi ∈ R +

n n
x x xn x x + + ⋯ + xn x x
2
+
2
+ ⋯ xn
n n
1 2 1 2 2
⩽ 1 2
⋯ ⩽ ⩽

x x
1 1 1

1
+
2
+ ⋯ +
xn
2. General Cauchy-Schwarz inequality

For all ai bi , ∈ R and n Z ∈ ⩾2

∑i aibi ∑i n ai ∑i n bi
n 2

2 2
⩽

=1 =1 =1

Students are generally unallowed to quote the above inequalities, thus the expected solutions will not use
the above methods. However, if you can see that the inequality can be proven via the above methods and
are unsure of how to do it as expected, then it can come in clutch if you can quickly prove the method you
want to use, then refer to it to prove the inequality. Remember - clarity is what makes a proof good.

AM/GM Proof (Pólya's Proof)

Lemma 1 x R ex- x .
∀ ∈ :
1
⩾

[Proof Let f x
] ex- -x
( ) =
1

f x ex-x- - 1

f x e
′( ) = 1
1
″( ) =

TPs at f x ′( ex-
) = 0 ⟹ x 1
= 1 ⟹ = 1

Now f ″(1) =e 0
= 1 > 0

Thus f x has only one turning point; min at x

( ) = 1
 x R fx f
∴ ∀ ∈ : ( ) ⩾ (1) equality at x = 1

ex- -x 1

ex- x
⟹ ⩾ 0

∴
1
⩾ ⬜

Now let A
x x where n Z and xi R for i n
+ + ⋯ + xn
n
1 2
+ +
= ∈ ∈ 1 ⩽ ⩽

x- x x x xn - x
By Lemma 1, e A e A e A n 1
1
2
1

A A A
1 2
⩾ , ⩾ , ⋯ , ⩾

x x xn -n x x x
Now e n by multiplying the above
+ + ⋯ +

A
1 2
⋯

A
1 2

n ⩾

en-n x xAn xn ⩾
1 2 ⋯

A n x x xn
x x xn n x x xn ⬜
⩾ 1 2 ⋯

+ + ⋯ +

n
1 2
⟹ ⩾ 1 2
⋯

HM/GM Proof
1 1 1

x +
x + ⋯ +
xn by AM/GM inequality
n
2
1 1 1

n x x xn
1
⩾ × × ⋯ ×
1 2

n
xx xn
=

n
1 2
⋯

⩽
nxx ⋯ xn ⬜
⟺ 1 2
1 1 1

x 1
+
x 2
+ ⋯ +
xn

AM/QM Proof
x x xn x x xn by Cauchy-Schwarz inequality
⏠⏣⏣⏣ ⏡⏣⏣⏣ ⏢
2 2 2 2
( 1
⋅1+ 2
⋅1+ ⋯ + ⋅ 1) ⩽ + + ⋯ + 1 + 1 + ⋯ +1
1 2

n terms
= nx x 2
1
+
2
2
+ ⋯ + xn 2

( x x + + ⋯ + xn )
2
x x 2
+
2
+ ⋯ + xn 2

n n
1 2 1 2
⩽
2

x x + + ⋯ + xn x x 2
+
2
+ ⋯ + xn
2

n n
1 2 1 2
⟹ ⩽ ⬜
Cauchy-Schwarz Inequality Proof (Induction)
Lemma 2 a a b b R ab ab
∀ 1
, a a b b
2
, 1
, 2
∈ :| 1 1
+ 2 2
| ⩽
2
1
+
2
2
×
2
1
+
2
2

[ Proof Define two arbitrary vectors a aa and b bb separated by an angle

] =
1
=
1
𝜃

-
2 2

1 ⩽ cos 𝜃 ⩽ 1

-a b a b | || ab | ⩽ | || |cos 𝜃 ⩽ | || |

ab a b
⟺ | ⋅ | ⩽ | || |

ab ab
∴ | 1 1
+ a a b b ⬜
2 2
| ⩽
1
2
+
2
2
×
2
1
+
2
2

RTP n Z ∑ aibi ∑ ai ∑ bi ∑i aibi

n 2
n n n
Now ( ) ∀ ∈ : ⩽
2 2
⟺ ⩽ ∑ni ai 2
× ∑ni bi 2

i i i
⩾2
=1 =1
=1 =1 =1 =1

where ai bi R for i n
, ∈ 1 ⩽ ⩽ .

The base case n is true by Lemma 2. = 2

Assume true for n k Z s t ∑ ai bi

k
∑ki ai = ∈ . . ⩽
2
× ∑ki bi 2

i
⩾2
=1 =1
=1

For n = k + 1, ( RTP ) ∑i aibi

k+1

=1
⩽ ∑ki ai
+1

=1
2
× ∑ki bi +1

=1
2

RHS = ∑ki ai ak =1
∑ki bi bk
2
+
2
+1
×
=1
2
+
2
+1

⩾ ∑ki ai ∑ki bi ak bk
=1
2
×
=1
2
+ +1 +1
by Lemma 2

⩾ ∑i k aibi ak bk
=1
+ +1 +1
by assumption

⩾ ∑i aibi ak bk
k
=1
+ +1 +1
( ∵ | x x | ⩾ )

= ∑ki aibi
+1

=1

∴ RHS LHS ⩾ Thus n = k true implies n k = +1 true

By the process of mathematical induction, the statement is proven true ∀ n Z
∈ ⩾2
⬜
Mean Chain & Cauchy Examples
This section presents various questions which are made simpler through the applications of the above
proven inequalities.
Mean Value Theorem (Proof):
(RTP If f x is continuous and differentiable on
) ( ) a b then there exists a tangent within the interval parallel
[ , ]

to the secant passing through the points a f a ( , ( )), ( , bfb ( ))

i.e. c a b f c f bb-a
∃ ∈ ( ,
-f a ): ′( ) =
( ) ( )

Let g x be the secant line passing through a f a and b f b By the point gradient formula,
( ) ( , ( )) ( , ( )).

g x -f a f bb-a
( )
-f a x-a g x f b -f a x-a f a
( ) =
( )

b-a
( )
( ) ⟹ ( ) =
( ) ( )
( )+ ( )

Let h x be the distance between the function and the secant line i.e. h x f x -g x
( ) ( ) = ( ) ( )

h x f x - f bb-a
( ) =
-f a x-a -f a
( )
( ) ( )
( ) ( )

Since h a h b then by Rolle's theorem c a b h c

( ) = ( ) = 0, ∃ ∈ ( , ): ′( ) = 0

h x f x - f bb-a
′( ) =
-f a ′( )
( ) ( )

Thus h c f c - f b -f a ( ) ( )

b-a
′( ) = ′( ) = 0

fc f b -f a ( ) ( )
⟹
b-a
′( ) = ⬜

Mean Value Theorem (integral form):

f c f b -f a
b
∫ f x dx (this is the mean value of a function)
( ) ( ) 1

b-a
′( ) =
a =
b-a ′( )

Let g x f x ( ) = ′( ).

b
⟹ ∃ c a b g c b-a ∫a g x dx
∈ ( , ): ( ) =
1
( ) where g x is continuous on
( ) ab
[ , ] and differentiable on ab
( , )

Geometric Interpretation: There exists some c in the interval ( ,ab ) where the area of rectangle ( b-a f c
) ( )

is equal to the area under the curve f x from a to b. ( )