TASK MATHEMATICS for ECONOMIC APPLICATIONS 16/03/2017 I M 1) Using the definition of a complex exponential we get:
D œ /"'3 œ / † /'3 œ /cos ' 3 sin 'œ /cos' 3sin'œ œ /cos' 3/sin' , the algebraic form of D.
Since # z 'ß #)1 the number , in the complex plane, has this position:D
I M 2) To achieve the requested matrix firstly we calculate its characteristic polynomial:
: œ œ " $ $ œ " $ $ œ
# # $ " ! "
" # % " # %
- -ˆ - -
- - -
- -
œ " $ $ " " $ œ
# % " #
- - - -
-
œ " - - # ( ' " - -œ " - - "- (. The three eigenvalues of are -" œ -# œ ", -$ œ (.
Since " †ˆœ " # $ " †ˆ
" # $
" # $
it follows 7 œ $ "1 Rank œ # and so the matrix is a diagonalizable one.
The second step is to find for every eigenvalue the corresponding eigenvectors.
For -"ß# œ " we solve the system:
- ˆ †—œÊ † œ Ê Ê
" # $ B ! B #B $B œ !
" # $ B ! B #B $B œ !
" # $ B ! B #B $B œ !
"
" " # $
# " # $
$ " # $
Ê B œ #B $B Ê" # $ —œ #B $B ß B ß B Þ # $ # $ Two linearly independent ei- genvectors corresponding to - œ " are —" œ #ß "ß ! and —# œ $ß !ß " . For -$ œ ( we solve the system:
- ˆ †—œÊ † œ Ê
& # $ B !
" % $ B !
" # $ B !
$
"
#
$
Ê Ê Ê
&B #B $B œ ! "#B "#B œ !
B %B $B œ ! 'B 'B œ !
#B $B œ ! œ #B $B
œ B œ B
" " # # $ $ # # $$
" # $ " # $
" $
# $
B B
B
B and so the ei- genvector corresponding to - œ ( is —$ œ Bß Bß B Ê —$ œ "ß "ß " Þ
So a matrix satisfying † œ ƒ† is œ .
# $ "
" ! "
! " "
I M 3) First method, by the transpose of the adjoint matrix divided by the determinant.
œ œ œ # † " † œ "# & œ (
# # $ # # $
" $ $ ! " "
" # % " # %
" " # $
# % " " .
(note that œ " † " † (, the product of the eingenvalues of the matrix). And so:
E.4 œ œ
$ $ " $ " $
# % " % " #
# $ # $ # #
# % " % " #
# $ # $ # #
$ $ " $ " $
' " "
# & #
$ $ %
.
" X
œ " E.4 œ " œ
(
' # $ 'Î( #Î( $Î(
" & $ "Î( &Î( $Î(
" # % "Î( #Î( %Î(
.
Second method, by elementary operations on >2/ 638/= 90 >2/matrix.
# # $ l " ! ! # # $ l " ! !
" $ $ l ! " ! ! # $Î# l "Î# " !
" # % l ! ! " ! " &Î# l "Î# ! "
´ ´
V V
V V V V
# " "
#
$ " "
# $ " #
#
# # $ l " ! !
! # $Î# l "Î# " !
! ! (Î% l "Î% "Î# "
´
"
# "
"
# #
%
( $
V V V
" " $Î# l "Î# ! !
! " $Î% l "Î% "Î# !
! ! " l "Î( #Î( %Î(
´
V V V V
" #
# $ $
%
" ! $Î% l $Î% "Î# !
! " ! l "Î( &Î( $Î(
! ! " l "Î( #Î( %Î(
V V"´$ $
%
" ! ! l 'Î( #Î( $Î(
! " ! l "Î( &Î( $Î(
! ! " l "Î( #Î( %Î(
.
In the right block we have the requested inverse matrix.
I M 4) If œ — —
+ + +
+ + +
+ + +
"" "# "$
#" ## #$
$" $# $$
" #
, since and belong to the Kernel, we have:
† œÊ œ Ê Ê
! ! œ !
" ! œ !
" ! œ !
—"
"" "# "$ "# "$
#" ## #$ ## #$
$" $# $$ $# $$
+ + + + +
+ + + + +
+ + + + +
Ê Ê
œ
œ
œ
+ + + + +
+ + + + +
+ + + + +
œ
"$ "# "" "# "#
#$ ## #" ## ##
$$ $# $" $# $#
;
† œÊ œ Ê Ê
" ! œ !
" ! œ !
! ! œ !
—#
"" "# "# "" "#
#" ## ## #" ##
$" $# $# $" $#
+ + + + +
+ + + + +
+ + + + +
Ê
œ
œ
œ
+ + + + +
+ + + + +
+ + + + +
œ
"" "# "# "# "#
#" ## ## ## ##
$" $# $# $# $#
,
.
Finally, from 0 "ß "ß " œ #ß "ß " , must satisfy:
" # œ # œ #
" " œ " œ "
" " œ "
œ Ê Ê
+ + + + +
+ + + + +
+ + + +
"# "# "# "# "#
## ## ## ## ##
$# $# $# $#
+$# œ "
.
So œ
#
"
" " "
# #
" "
.
To find a basis for the Kernel and a basis for the Image, since Rank œ " we have simply to note that —" and —# are linear independent, so UO/< 0 œ — —"ß # and
UM77 0 œ#ß "ß ".
II M 1) The equation 0 Bß Cß D œ B/ CD C/BD D/BC œ ! verifies 0 "ß "ß ! œ ! . f0 œ / CD C/BD D/BCà B/CD /BD D/BCà B/CD C/BD /BC from which we get f0 "ß "ß ! œ !à !à / #. Since 0 T Á !Dw , the equation defines an impli- cit function D œ D Bß C with:
fD "ß " œ 0 T à œ !à ! œ !ß !
0 T 0 T / /
0 T
Bw
w w # #
D D
Cw .
II M 2) The problem Max min
u c Î 0 Bß C œ B BC is equivalent to the problem:
Þ Þ %B C Ÿ %
# #
# #
Max min
u c Î 0 Bß C œ B BC ; the feasible region is drawn in red in the figure Þ Þ %B C % Ÿ !
# #
# # on
the next page; the objective function is continous and the feasible region is a bounded and closed set, so by Weiertrass Theorem the problem has absolute maximum and mini- mum. The Lagrangian function of the problem is:
ABß Cß-œB BC# #-%B C %# #
whose gradient vector is f œ #A B C# )-B #BCà #-Cà %B C %# # . KUNH-TUCKER CONDITIONS
First case (free optimization): , the unique solution
- œ ! -
# œ ! œ !
œ ! œ !
Ÿ ! Ê
œ !
Ÿ
B C B
#BC C
%B C % ! %
#
# #
point is !ß ! . # !ß ! ! !ß !
# !
‡ 0 # C ‡ # ‡
C #B !
œ , œ with œ !; second order conditions are useless to give us informations on the nature of the point !ß ! , but from the figure on the next page it is easy to see that !ß ! is a saddle point for the obje- ctive function of the problem.
Second case (active constraint):
- -
- -
- -
Á ! Á !
# ) œ ! # ) œ !
# œ ! œ !
œ ! œ
B C B Ê B C B
#BC C #C B
%B C % %B C %
# #
# # # # from which we get:
-
-
- - Á !
# % œ !
œ ! œ
Ê
Á ! œ "Î%
œ ! œ B "
C
%B %
C B „"
#
and
-
- -
-
- -
-
- -
-
- -
Á ! œ ) œ
"# œ Ê
Á ! œ ) œ
# " $ # œ
C #
B
# %
C #
B
!
# #
#
# #
from which we get:
- -
- -
- -
- -
Á ! Á !
œ $ œ #!Î*
œ "Î# œ #Î$
œ "Î# œ #Î$
Ê Ê
Á ! Á !
œ $ œ #!Î$
œ "Î# œ #Î$
œ "Î# œ #Î$
C C
B B
C „ C „
B B
# and # .
So we get six constrained critical points:
T"ß# œ„"ß ! T, $ß% œ #Î$ß „#!Î$, Q +B ? ; T&ß' œ "Î#ß„$, 738 ? . 0T"ß#œ ", 0T$ß%œ %%Î*, 0T&ß'œ &Î%.
Q +B 0 œ %%Î* at points T à 738 œ &Î%$ß% 0 at points T&ß'.
Note also that by figure below point is a local maximum; is not a maximum point.T# T"
On the figure below there are drawn zero level curve (yellow), positive level curves (blue) and negative level curves (pink).
II M 3) For 0 Bß Cß D œ logB C / DBB C CD# $, differentiable function, we have: H 0 T@ ! œf0 T! † @, where is the unit vector of @ "ß "ß " . T œ "ß !ß "!
f0 Bß Cß D œ " / #BC à " B D à / $CD
B C B C
DB # $ DB #
f0"ß !ß " œ " " ! à " " " à " ! œ #à "à " Þ Since
@ œ " à " à " H 0 T œ #à "à " † " à " à " œ !
$ $ $ $ $ $
we get @ ! .
II M 4) The Mac Laurin polynomial of degree 2 at point !ß ! for the function is0
T Bß C œ 0 !ß ! f0 !ß ! " !ß ! 0 !ß ! œ !
# † Bß C #Bß C † † B
‡0 C ; ,
f0 œ / B C B C / B C à /B C B C / B C œ
f0 œ" B C / B C à " B C / B C à f0 !ß ! œ "ß " ;
‡0 Bß C œ / " B C / / " B C / œ
/ " B C / / " B C /
B C B C B C B C
B C B C B C B C
‡0 Bß C œ ‡0 œ
# B C / B C / # !
B C / # B C / !ß !
! #
B C B C
B C B C
, .
So T#Bß C œ B C " # # œB C
# B C# # B C# #.