Sie sind auf Seite 1von 10

No part of this publication may be reproduced or distributed in any form or any means, electronic, mechanical,

photocopying, or otherwise without the prior permission of the author.

GATE SOLVED PAPER


Mathematics
Numerical Analysis

Copyright © By NODIA & COMPANY

Information contained in this book has been obtained by authors, from sources believes to be reliable. However,
neither Nodia nor its authors guarantee the accuracy or completeness of any information herein, and Nodia nor its
authors shall be responsible for any error, omissions, or damages arising out of use of this information. This book
is published with the understanding that Nodia and its authors are supplying information but are not attempting
to render engineering or other professional services.

NODIA AND COMPANY


B-8, Dhanshree Tower Ist, Central Spine, Vidyadhar Nagar, Jaipur 302039
Ph : +91 - 141 - 2101150
www.nodia.co.in
email : enquiry@nodia.co.in
GATE SOLVED PAPER - MA
NUMERICAL ANALYSIS

YEAR 2005 ONE MARK

Q. 1 An iterative method to find the n th root ^n d N h of a positive number a is given


by

xk + 1 = 1 :xk + x n − 1 D
a
2 k

A value of n for which this iterative method fails to converge is

Y
(A) 1 (B) 2

N
(C) 3 (D) 8

Q. 2

PA
Suppose the function u ^x h interpolates f ^x h at x 0, x1, x2, ..., xn − 1 and the function
v ^x h interpolates f ^x h at x1, x2, ..., xn − 1, xn . Then a function F ^x h which interpolates
f ^x h at all the points x 0, x1, x2, ..., xn − 1, xn is given by
^xn − x h u ^x h − ^x − x 0h v ^x h
M ^xn − x h u ^x h + ^x − x 0h v ^x h

O
(A) F ^x h = (B) F ^x h =
^xn − x 0h ^xn − x 0h
(C) F ^x h =
^xn − x 0h C
^xn − x h v ^x h + ^x − x 0h u ^x h
(D) F ^x h =
^xn − x h v ^x h − ^x − x 0h u ^x h
^xn − x 0h

&
IA
YEAR 2005 TWO MARKS

D
Q. 3 Suppose the iterates xn generated by
2f ^xn h
O
xn + 1 = xn −
f ' ^xn h

N
where f ' denotes the derivative of f , converges to a double zero x = a of f ^x h.

©
Then the convergence has order
(A) 1 (B) 2
(C) 3 (D) 1.6
J 2 α − 1N
K O
Q. 4 Suppose the matrix M = K α 2 1 O has a unique Cholesky decomposition of the
K− 1 1 4 O
L P
form M = LL , where L is a lower triangular matrix. The range of values of α is
T

(A) − 2 < α < 2 (B) α > 2


(C) − 2 < α < 3/2 (D) 3/2 < α < 2

Q. 5 The Runge-Kutta method of order four is used to solve the differential equation
= f ^x h, y ^0 h = 0
dy
dx
with step size h . The solution at x = h is given by
(A) y ^h h = h ;f ^0 h + 4f b h l + f ^h hE (B) y ^h h = h ;f ^0 h + 2f b h l + f ^h hE
6 2 6 2
(C) y ^h h = h 6f ^0 h + f ^h h@ (D) y ^h h = h ;2f ^0 h + f b h l + 2f ^h hE
6 6 2
GATE SOLVED PAPER - MA NUMERICAL ANALYSIS

Q. 6 The values of the constants α, β, x1 for which the quadrature formula


1

# f^x hdx = αf ^0 h + βf ^x1h


0

is exact for polynomials of degree as high as possible, are


(A) α = 2 , β = 1 , x1 = 3 (B) α = 3 , β = 1 , x1 = 2
3 4 4 4 4 3
(C) α = 1 , β = 3 , x1 = 2 (D) α = 2 , β = 3 , x1 = 1
4 4 3 3 4 4

Q. 7 In solving the ordinary differential equation y' = 2x , y ^0 h = 0 using Euler’s


method, the Iterates yn , n d N satisfy

Y
(A) yn = 2x n2 (B) yn = 2xn

N
(C) yn = xn xn − 1 (D) yn = xn − 1 + xn

PA
YEAR 2004 ONE MARK

M
Q. 8 The equation x6 − x − 1 = 0 has

O
(A) No positive real roots (B) Exactly one positive real root

C
(C) Exactly two positive real roots (D) All positive real roots

An iterative scheme is given by xn + 1 = 1 c16 − 12 m, n d N , " 0 , . Such a scheme,


&
Q. 9
5 xn
with suitable x 0 , will

IA
(A) Not converege (B) Converge to 1.6
(C) Converege to 1.8 (D) Converge to 2

D
YEAR 2004
O TWO MARKS

Q. 10
N
To find the positive square root of a > 0 by solving x − a = 0 by the Newton-
2

©
Raphson method, if xn denotes the n th iterate with x 0 > 0 , x 0 =
Y a , then the
sequence "xn, n $ 1, is
(A) Strictly decreasing (B) Strictly increasing
(C) Constant (D) Not convergent

YEAR 2003 ONE MARK

Q. 11 Let M be the length of the initial interval 6a 0, b 0@ containing a solution of f ^x h = 0 .


Let 6x 0, x1, x2, ...@ represent the successive points generated by the bisection method.
Then the minimum number of iterations required to guarantee an approximation
to the solution with an accuracy of ε is given by
ε ε
log a M k log a M k
(A) − 2 − (B) − 2 +
log 2 log 2
ε
log ^Mεh log a M k
(C) − 2 + (D) − 2 −
log 2 ^log 2h2
GATE SOLVED PAPER - MA NUMERICAL ANALYSIS

2 2
Q. 12 On evaluating # # ^x +1 y h dxdy numerically by Trapezoidal rule one would get
the value 1 1

(A) 17 (B) 11
48 48
(C) 21 (D) 17
48 52

YEAR 2003 TWO MARKS

Q. 13 If the scheme corresponding to the Newton-Raphson method for solving the system
of nonlinear equation : x2 + 2 − 10 = 0 , x2 y − 3 = 0 is xk + 1 = xk + " f ^x, y h,^x , y h and k k

yk + 1 = yk + "g ^x, y h,^x , y h then f ^x, y h and g ^x, y h are respectively given by
k k

^x2 + y2 − 10h ^x2 y − 3h


(A) − and −

Y
2x x2
x2 ^y2 − x2 + 10h − 6y

N
y3 − 10y + 3
(B) and
2x ^x − y h

A
2 2
x2 − y2
x2 ^3y2 − x2 − 10p + 6y h

P
y3 − 10y + 3
(C) and
2x ^x − y h
2 2
x2 − y2
(D) −^x2 − y2 − 10h and −^x2 y − 3h
M
O
A lower bound on the polynomial interpolation error e2 ^x h for f ^x h = ln ^x h, with

C
Q. 14

x 0 = 2 , x1 = 2 , x2 = 4 and x = 5 is given by

&
4
(A) 1 (B) 1
256 64
(C) 1
IA (D) 0

D
512
Q. 15 Consider the Quadrature formula

O
h

f ^x h dx = 'αf ^0 h + βf b 3h l + γf ^h h1 h
N
# 4
0

©
The values of α , β , γ for which this is exact for polynomials of as high degree as
possible, are
(A) α = 5 , β = 8 , γ =− 1 (B) α = 1 , β =− 1 , γ = 3
18 9 6 2 4 4
(C) α = 0, β = 1, γ =− 1 (D) α = 1, β = 2, γ = 3
4
Q. 16 Consider the (Cholesky’s) algorithm given below for LLT decomposition of a
symmetric Positive Definite matrix A
Compute L11 = A 11 1/2

For i = 2 to N
A
Compute Li, 1 = i, 1
L i, 1
For i = 2 to N
j−1 1/2
Compute L j, j = f A j, j − / L2j,m p
m=1

Right alternative for filling the shaded box to complete the above algorithm is
GATE SOLVED PAPER - MA NUMERICAL ANALYSIS

For i = j + 1 to N
j−1
(A) Compute L = 1 f A − L i, m L j, m p
i, j
L j, j i, j /
m=1

For i = j to N
j−1
(B) Compute L = 1 f A − L i, m L j, m p
i, j
L j, j i, j /
m=1

For i = j to N
j−1
(C) Compute L = 1 f A − L i, m L j, m p
i, j
L j, j i, j m/
=1

For i = j + 1 to N

Y
j−1
(D) Compute L = 1 f A − L2i, m p
i, j
L j, j i, j /
N
m=1

PA
YEAR 2002 ONE MARK

Q. 17 If f ^x h has an isolated zero of multiplicity 3 at x = ξ , and the iteration


3f ^xn h
M
O
xn + 1 = xn − , n = 0, 1, 2, ...
f ' ^xn h

C
converges to ξ , then the rate of convergence is
(A) Linear

&
(B) Faster than linear but slower than quadratic

IA
(C) Quadratic
(D) Cubic

Q. 18
D
The best possible error estimate in the Gauss-Hermite formula with 3 points, for

O
3

#xe
N
4 −x2
calculating the integral dx is
(A) 0 −3 (B) 0.30

©
(C) 0.65 (D) 1.20

YEAR 2002 TWO MARKS

Q. 19 The fourth divided difference of the polynomial 3x3 + 11x2 + 5x + 11 over the
points x = 0, 1, 4, 6 and 7 is
(A) 18 (B) 11
(C) 3 (D) 0.059

Q. 20 The polynomial of least degree interpolating the data ^0, 4h, ^1, 5h, ^2, 8h, ^3, 13h is
(A) 4 (B) 3
(C) 2 (D) 1

Q. 21 For the matrix


R V
S0 1 − 1 1W
S1 0 0 2W
S W
S− 1 0 2 0W
S1 2 0 0W
the bound for Tthe X
eigen values predicted by Gershgorin’s theorem is
(A) 3 (B) 1
(C) 2 (D) 4
GATE SOLVED PAPER - MA NUMERICAL ANALYSIS

YEAR 2001 ONE MARK

Q. 22 If T and d are the forward and the backward difference operators respectively,
then T − d is equal to
(A) − Td (B) Td
(C) T + d (D) T
d
Q. 23 One root of the equation ex − 3x2 = 0 lies in the interval ^3, 4h, The least number
of iterations of the bisection method so that error # 10−3 are
(A) 10 (B) 8
(C) 6 (D) 4

YEAR 2001 TWO MARKS

The least squares approximation of first degree to the function f ^x h = sin x over

Y
Q. 24
π π
the interval 9− 2 , 2 C is

N
A
(A) 243x (B) 242x

P
π π
24 x

M
(C) (D) 24π
π

Q. 25
O
The order of the numerical differentiation formula
f m^x 0h = 1 2 >
C
− " f ^x 0 − 2h h + f ^x 0 + 2h h,
12h + 16 " f ^x 0 − h h + f ^x 0 + h h, − 30f ^x 0hH
is
&
IA
(A) 2 (B) 3

D
(C) 4 (D) 1

O
Q. 26 The method

N
yn + 1 = yn + 1 ^k1 + 3k2h, n = 0, 1, ...
4

©
k1 = hf ^xn, yn h
k2 = hf bxn + 2h , yn + 2 k1 l
3 3
is used to solve the initial value problem
y' = f ^x, y h =− 10y, y ^0 h = 1
The method will produce stable results if the step size h satisfies
(A) 0.2 < h < 0.5 (B) 0 < h < 0.5
(C) 0 < h < 1 (D) 0 < h < 0.2

YEAR 2000 TWO MARKS

Q. 27 The smallest value of x ^ x < 1h correct to two decimal places satisfying the
equation
3 5 7 9 11
x − x + x − x + x − x + ... = 0.4431135
3 10 42 216 1320
is
(A) 0.58 (B) 0.47
(C) 0.44 (D) 0.88
GATE SOLVED PAPER - MA NUMERICAL ANALYSIS

Q. 28 The Jacobi’s iteration method for the set of equations


x1 + ax2 = 2 ,
2ax1 + x2 = 7 , da =
Y 1 n converges for
2
(A) All values of a (B) a = 1
(C) a < 1 (D) 1 < a < 3
2 2 2

Q. 29 The interpolation polynomial of highest degree which corresponds the functional


values f ^− 1h = 9 , f ^0 h = 5 , f ^2 h = 3 , f ^5 h = 15 , is
(A) x3 + x2 + 2x + 5 (B) x2 − 3x + 5

Y
(C) x 4 + 4x3 + 5x2 + 5 (D) x + 5

N
PA
YEAR 1999 TWO MARKS

Q. 30 The maximum step size h such that the error in linear interpolation for the
function y = sin x in 60, π@ is less than 5 # 10−5 is
M
O
(A) 0.02 (B) 0.002

C
(C) 0.04 (D) 0.06

Q. 31 The second order Runge-Kutta method is applied to the initial value problem

(A) y 0 ^h − 1h2 &


y' =− y , y ^0 h = y 0 with step size h . Then y ^h h is
(B) 0 ^h2 − 2h + 2h

IA
y
2

D
(C) 0 ^h2 − 2h + 2h (D) y 0 c1 − h + h + h m
2 3
y
6 2 6

O
N
The iterative scheme xn + 1 = xn b1 + x 2 l converges to a . The convergence is
a
Q. 32
2 n

©
(A) Linear (B) Quadratic
(C) Cubic (D) Biquadratic

YEAR 1998 ONE MARK

Q. 33 The sequence "xn , of m # m matrices defined by the iterations


Xn + 1 = 2Xn − Xn , AXn , n = 0, 1, 2
when X 0 = 1, the identity matrix converges to A−1 , if and only if each eigen value
λ of A satisfies
(A) λ < 1 (B) λ − 1 < 1
(C) λ + 1 < 1 (D) None of the above

Q. 34 A nontrivial solutions of
x2 y m + xy' + 4y = 0 , x > 0 are
(A) Bounded and non periodic (B) Unbounded and non-periodic
(C) Bounded and periodic (D) Unbounded and periodic

Q. 35 Every solution of y m + ay' + by = 0 , where a and b are constants, approximate to


zero as x " 3 provided
(A) a > 0, b > 0 (B) a > 0, b < 0
(C) a < 0, b < 0 (D) a < 0, b > 0
GATE SOLVED PAPER - MA NUMERICAL ANALYSIS

Q. 36 The boundary value problem


y m + ky = 0 , y' ^0 h = 0 , y' ^πh = 0
has non trivial solution for
(A) All negative value of k (B) All the values of k
(C) k = 0 (D) k = ! n, n = 1, 2, ...,

YEAR 1998 TWO MARKS

Q. 37 The Lagrange interpolation of the function f ^x h = x2 on the interval 6− 1, 1@ has


the least maximum error, if the nodes of interpolation are
(A) zeros of the Legender polynomials, Pn
(B) zeros of Chebyshev polynomials, In
(C) #k/ ^n − 1h : 0 # k # n − 1-

Y
(D) "k/ ^n + 1h : 1 # k # n ,

Q. 38

N
The following matrix admits a cholesky (also called LL*) decomposition :

A
1 i 1 2i
(A) > H (B) > H

P
i 1 − 2i 1

M
1 −2 1 1
(C) > H (D) > H

O
−2 5 2 2

YEAR 1997 C TWO MARKS

Q. 39
&
The Hermite interpolating polynomial for the function f ^x h = x6 based on − 1, 0

IA
and 1 is
(A) x 4 − 2x2 (B) 2x 4 − x2
(C) x 4 + 2x2
D (D) 2x 4 + x

Q. 40
O
The system of equations

N 3x + 2y = 4.5

©
2x + 3y = 5.0
− y + 2z =− 0.5
is to be solved by successive over relaxation method. The optimal relaxation
factor ω opt , rounded upto two decimal places is given by
(A) 1.23 (B) 0.78
(C) 1.56 (D) 0.63

YEAR 1996 TWO MARKS

Q. 41 The equation ex − 4x2 = 0 has a root between 4 and 5, fixed point iteration with
iteration function 1 ex/2 :
2
(A) Diverges (B) Converges
(C) Oscillates (D) Converges monotonically

Q. 42 If x = ξ is a double root of the equation f ^x h = 0 and if f m^x h is continuous in a


neighbourhood of ξ then the iteration scheme for determining ξ
2f ^xn h
xn + 1 = xn −
f ' ^xn h
GATE SOLVED PAPER - MA NUMERICAL ANALYSIS

has the order


(A) 2 (B) 1
(C) Less than 2 (D) Less than 1

Q. 43 The formula A 0 f b− 1 l + A2 f ^0 h + A2 f b 1 l which approximates the integral


1 2 2
# f^x hdx is exact for polynomials of degree less than or equal to 2 if
−1

(A) A 0 = A2 = 4 , A1 = 2 (B) A 0 = A1 = A2 = 1
3 3
(C) A 0 = A2 = 4 , A1 =− 2 (D) None of the above

Y
3 3

N
PA
YEAR 2000 THREE MARKS
π
2

M
Q. 44 The value of # 1 − 0.162 sin2 φdφ by Simpson’s one-third rule by taking the

O
0

step sizes as π

C
12
(A) 1.5058 (B) 1.5759
(C) 2.5056
& (D) 1.5056

YEAR 1997
IA THREE MARKS

D
O
Q. 45 The range of the absolute stability of the multistep method
yn − 1 = yn + h ^23fn − 16fn − 1 + 5fn − 2h

N
12
for the differential equation y' = f ^x, y h is

©
(A) b− 6 , 0 l
11
(B) b− 11, 0 l
6
(C) b 0, − 6 l (D) b 0, − 11 l
11 6

**********
GATE SOLVED PAPER - MA NUMERICAL ANALYSIS

ANSWER KEY
NUMERICAL ANALYSIS
1 2 3 4 5 6 7 8 9 10
(A) (D) (D) (C) (A) (C) (C) (B) (D) (B)
11 12 13 14 15 16 17 18 19 20
(A) (A) (A) (C) (A) (D) (A) (C) (D) (C)
21 22 23 24 25 26 27 28 29 30
(A) (B) (D) (B) (A) (D) (B) (C) (B) (A)
31 32 33 34 35 36 37 38 39 40
(B) (B) (A) (A) (A) (B) (B) (C) (B) (A)
41 42 43 44 45

Y
(A) (A) (C) (D) (A)

N
PA
M
O
C
&
IA
D
O
N
©

Das könnte Ihnen auch gefallen